Logan Ford

Teddy Lee

When Covid-19 first hit, many companies had to rapidly reorient their operations – from altering their supply chains to implementing social distancing and work-from-home orders.
One sector that was particularly hard hit was the restaurant industry. When in-person dining was suspended, restaurants turned to delivery to survive. But setting up a delivery service from scratch takes time and money that many establishments, with their income drying up, did not have. Suddenly, food delivery platforms became an economic lifeline.
To ensure restaurants could onboard quickly and seamlessly in these critical early days, one of the leading delivery platforms turned to Scale.
<h3 id="scaling-labeling">Scaling labeling</h3>
Every time a restaurant joins one of the delivery platforms or changes its menu, the menu’s contents need to be inputted into the platform so users can select their food. Inputting menus manually is slow, expensive, and leads to mistakes. With restaurants relying on delivery to reach customers and support their staff, a leading delivery platform knew it was vital to ensure an efficient process for its partner restaurants.
This presented Scale with unique latency and quality control challenges, requiring both an incredibly high level of accuracy with a fast turnaround to provide operational efficiencies at a critical time for our customer. Our aim was to use the combination of partial automation and human quality control in our <a href="https://scale.com/document" target="_blank">Scale Document</a> labeling pipeline to make dramatic efficiency improvements over manual processes. In all, we nearly halved the time it takes our customer to process menus and reduced critical error rates to &lt;1% for all items labeled, allowing them to onboard restaurants smoothly exactly when they needed it.
<h3 id="building-the-right-data-infrastructure">Building the right data infrastructure</h3>
While processing standardized documents such as food menus, government IDs or loan application forms might be intuitive and simple for humans, it is a surprisingly nontrivial problem for algorithms, involving much more than automatic transcription. In particular, the data often contains many dependencies that need to be captured accurately. For example, ordering a pepperoni pizza doesn’t just mean choosing the type, but also the size, the type of base, choosing to substitute or customize toppings (which can also vary in price depending on the size of the pizza, creating more interdependencies), and whether to add extras like burgers or sandwiches. It’s important to capture these dependencies in the menu correctly, otherwise inputting errors risk causing losses for both the restaurant and the delivery platform.

DD_decision

Algorithmically, this has to be represented as complex decision trees made up of categories, items, and lists of options, capturing the relationships between them. Thankfully, both our document processing pipeline and form capture tools were already set up to capture these types of nested data structures automatically without the need for customization. This meant we were able to quickly pivot our tools for our customer’s requirements.

### Judicious Automation and Scaling
Once we have the right tooling in place to pull the raw data and label complex data structures, we can then start processing the data. Given how many menus delivery platforms were processing at the start of shelter-in-place, and how important for restaurants’ continued operations it was to process this data quickly with no drop-off in quality, we need to make this process as efficient as possible. 

At Scale AI, we believe that a mix of judicious labeling automation with continued human oversight is the only way to provide data at the scale, quality, and low cost needed to enable many enterprise applications of AI. In particular, we aim to automate the lower-skilled parts of the data labeling pipeline to focus human review on the hard parts – quality control, edge cases, and the most complex data types. That way, we can help guarantee quality without sacrificing efficiency.

We use automation to prefill menu taxonomies for expert human labelers to review and confirm. 

<figure class="expand-image">
<video autoplay muted loop>
 <source src="https://videos.ctfassets.net/1qgw036lpkfx/3frjAvFdxn4pE0ZHJBnlaO/878291c892968b416019b22a4ecf62e8/scale-document-blog-video.mp4" type="video/mp4" />
</video>
 
<figcaption>We trained our form capture tool to automatically understand document structure and predict the next transcription.</figcaption>
</figure>

And when our labelers needed to add inputs manually, we built a “smart suggestion” feature that provides them with ready prompts for common items that they have already encountered in their batch of menus. 

Both have helped improve both the efficiency and accuracy of labeling simultaneously: allowing us to transcribe more than 3,000 menus in a single day and return menus within short turnaround times to onboard restaurants within 24 hours from start-to-finish. By drawing on our wide network of well-trained labelers and augmenting their work with careful automation, we’re able to handle changes in demand dynamically – essential for providing a smooth service during sudden peaks in demand.

### Custom Benchmarks and Integrations
Once data is flowing smoothly and efficiently through this labeling pipeline, the next step is to ensure that the data is being labeled to an incredibly high standard, with almost zero errors. We’ve developed a range of techniques to guarantee label quality, including the randomized use of benchmark quality tests, confidence-based consensus, and, for ambiguous labeling requirements, our own [automated benchmark generation system](https://scale.com/blog/how-to-label-1m-data-points-week). 

Benchmarking is not always an objective science – the most effective benchmarks are often task-specific. The high stakes of onboarding restaurants during a pandemic required particularly high guarantees of labeling quality. To meet them, we worked closely with our customer to build custom benchmarks for this workflow, turning the most complex mislabeled menus into benchmarks that assessed our labelers’ performance.

In surveying mislabeled menus, we noticed two key groups of errors. One was associated with the cross-referencing of menu sections between many items and mistaking how they relate to each other, while the other group tended to be mistakes from larger menus where it may be less obvious when certain features of the menu are missing. By identifying these key groups and creating benchmarks out of the associated menus, we can ensure that tasker performance in such cases exceeds the necessary standard to meet menu quality needs.


Menu_dd_photos

### What’s Next
AI models are rapidly becoming critical infrastructure for a huge range of businesses – a trend of automation that the COVID pandemic has rapidly accelerated. But gathering effective, accurate, and unbiased data currently remains such a challenging task that it is preventing real-world AI systems from reaching their full potential and presenting high complexity barriers for smaller companies.

We’re tackling a whole host of problems to help accelerate that progress. Right now, teams at Scale AI are making better use of machine learning to make labeling orders of magnitude more efficient, building tools to process increasingly complex data types, such as 3D point clouds in computer vision, and developing new infrastructure tools to help streamline the management of data. Our first such management tool, [Scale Nucleus](https://scale.com/blog/introducing-scale-nucleus), is already helping our computer vision customers automate time-consuming manual steps in the ML development process. We’re now working on deploying it for our customers in natural language. 

Ultimately, we want to make it as easy to deploy AI as any other type of software. If you’re interested in joining us in solving these problems, take a look at our [careers page](https://scale.com/careers) for our latest open positions. If you have projects that require high-quality data labeling, [let our team know](https://scale.com/sales)!


How Scale AI worked with a leading food delivery platform to ensure restaurants could onboard quickly and seamlessly. 

Rapidly scaling menu transcription to support restaurants during the pandemic

Scaling Menu Transcription Tasks with Scale Document

Shannon Hong

The Scale AI team hosted its first ever Hackathon last month. The hackathon was designed to:
<ol>
<li>Encourage innovative thinking to solve some of Scale’s biggest challenges</li>
<li>Prototype and potentially launch impactful projects that make Scale better in practice</li>
<li>Facilitate long-term cross-functional relationships</li>
<li>Give all Scaliens a creative break from their daily work.</li>
</ol>
Scaliens across functions and teams leapt to the challenge, proposing 63 ideas and ultimately working on 10 separate projects comprising of 37 team members. Projects ranged from new product proposals to better internal search, from refactoring tech debt to streamlining instructions for Taskers.

Nofel_Naoman_Hackathon_Quote

## Project Highlights: 

The following are just some of the projects Scaliens tackled during the Hackathon. As a disclaimer, most of these projects are not live, nor do they indicate immediate plans to build these ideas.

### Super Instructions 

__Winner:__ Best In Show

__Team Members:__ [Shaun VanWeelden](https://www.linkedin.com/in/svanweelden/), [Taisia Karaseva](https://www.linkedin.com/in/taisiakaraseva/), [Jeff Lin](https://www.linkedin.com/in/jeffrey-linj/), [Galen Bertozzi](https://www.linkedin.com/in/galenbertozzi/), [Rashid Lasker](https://www.linkedin.com/in/rashidlasker/), [Ivan Roumec](https://www.linkedin.com/in/ivan-roumec-b0450a6b/), [Nofel Naoman](https://www.linkedin.com/in/nofel-naoman-aa28a0142/), [Lawder Paul](https://www.linkedin.com/in/nba/)

__Problem Statement:__
Annotation instructions are critical for the development of high-quality training data. Today’s annotation instructions are mostly made on static, non-interactive documents, not connected to the API payloads creating the task and requiring 3rd party solutions to host.

__Solution:__
The Super Instructions team tackled this problem in 3 ways:

  1. Extending the Task Schema used in our APIs to accommodate labeling         instructions.
  2. Creating an in-app instruction builder that simultaneously creates the     instructions as well as the API payloads that would need to be leveraged     in task creation.
  3. Bringing a world-class interactive design to the in-app educational         experience for those doing the annotating, including features like           search, navigation, and zoomable images. 

The goal of this prototype was to re-imagine the experience both our customers and Taskers have when creating and completing work on Scale’s platform. We firmly believe better instruction design and more consistency leads to higher-quality annotations for all involved. 

### Corp Search

__Winner:__ Audience Choice

__Team Members:__ [Lawder Paul](https://www.linkedin.com/in/nba/) and [James Lennon](https://www.linkedin.com/in/james-lennon-30421078/)

__Problem Statement:__ 
“Corp” is Scale AI’s admin portal for employees to gather information on on-going projects and to influence the quality and outcome of data annotation. The Scale AI team has launched countless tools and introduced new functionalities to Corp to improve our operations but the Corp portal itself was not very user friendly. There are now so many tools in Corp that it can be hard to find the exact tool, and for new Scaliens it’s not always clear what different tools do. 


Scale_Corp_Search

Scale Corp Search

__Solution:__ 
Corp Search allows employees to quickly search across all areas and tools within Corp, packaged in an elegant "command-line" interface that can be surfaced anywhere in Corp.

### Corp Migration

__Winner:__ Most Immediately Impactful

__Team Members:__ [Diego Muracciole](https://www.linkedin.com/in/diegomura/), [Federico Marcos](https://www.linkedin.com/in/federicomarcos/), [Martin Cabral](https://www.linkedin.com/in/martin-cabral-b0897615/)

__Problem Statement:__ As stated in the Corp Search project, Corp had become a complex and sophisticated tool used by many to provide the best service to our customers. Many of the Tasker and customer-facing features and tools were tied together or interlinked with Corp, making additional engineering work more challenging. 

__Solution:__ 
The team spent the week separating these Tasker and customer-facing features were separated from Corp. The team also developed a framework to encourage internal modularity and code reuse to enable faster deployment in the future as Scale continues to grow. The team also ensured backwards compatibility with previous code to ensure daily operations could continue. This project has already enabled several refactors and performance improvements. 

### The Great Refactor

__Team Members:__ [Andrew Liu](https://www.linkedin.com/in/liuandrewk/), [Nick Induni](https://www.linkedin.com/in/nick-induni-44357292/), [Steven Hao](https://www.linkedin.com/in/steven-hao-160b9638/)

__Problem Statement:__ Like many fast-growing startups, Scale AI has accumulated some level of tech debt in order to quickly deliver new features and products to our customers. 

__Solution:__
The Great Refactor team looked to tackle this issue by investing a week of their time to improve code quality to optimize for long-term maintainability and development velocity.

### Tasker Scoresheets

__Team Members:__ [Akshat Bubna](https://www.linkedin.com/in/akshat-bubna-188885103/), [Rilka Li](https://www.linkedin.com/in/rilka/), [Alan Yu](https://www.linkedin.com/in/alan-yu-954926110/)

__Problem Statement:__ Although Taskers (our global data annotation workforce) regularly receive feedback on the quality of their submissions, it is difficult to provide granular feedback with specific areas of strength and weakness.

__Solution:__ 
The team developed Scoresheets that could be accessed on the Tasker dashboard. The dashboard allows Taskers to track their individual performance on specific skills over time and provides learning resources to improve at weak skills.

### Scale Grep

__Team Members:__ [Jonathan Fung](https://www.linkedin.com/in/jonfung1/)

__Problem Statement:__ Instead of searching through command-line grep, IDE search, or Github search, which can be slow and unwieldy, engineers need quick, interactive, on-the-fly code search that can generate quickly sharable links.

__Solution:__ 
We deployed two open-source projects, Livegrep and Webgrep internally, and added automated index-generation so that the code in our git repositories gets indexed every 2 hours. Now, engineers can search through our large codebase in milliseconds and click code fragments to generate sharable links to exact lines of code.

## Facilitating Interactions to Foster Cross-Functional Relationships

In addition to working on impactful projects, another goal of the Hackathon was to facilitate cross-functional relationships. Fostering these connections when everyone is remote can be particularly tricky. How do you recruit teammates to work on your project when you can’t organically run into folks in hallways? How do you whiteboard ideas, or learn a new skill from teammates when you can’t physically sit together?

### Pitch Session 

To help idea originators find teammates, we held a company-wide pitch session a week before the Hackathon. Scaliens then spent the following week signing up for the pitched projects and forming cross-functional teams. 

### Events and Activities


Scale_Hackathon_Zoom

To help Scaliens develop new skills, we organized workshops throughout the week of the Hackathon. Workshops included:

  - A session on Storybook for planning engineering deployments
  - A session on Figma to understand Scale’s design resources and to             enable better designing and prototyping
  - A session on Tableau to enhance data visualization. 

We also organized daily social activities and special events throughout the week to facilitate social bonding. These activities included a Daily “Brain Game” and a virtual Game Night. 

To help everyone get into the spirit of the Hackathon (whether they were on a team or not), we fully branded the hackathon experience. Our Brand Director, [Ty Wilkins](https://www.linkedin.com/in/tywilkins/), created custom Zoom backgrounds, and branded swag, while Product Designer, [Chun Jiang](https://www.linkedin.com/in/chun-jiang-0428/), designed our *Paranormal Hacktivity* mascot.


![scale_blog_hackathon_ghost](//images.ctfassets.net/1qgw036lpkfx/1MOMHZ4lAmvgsNDSxgZCFY/bc103016d8fc58881eba0d8d87b201b5/scale_blog_hackathon_ghost_04.svg)

### Presentations and Hackathon Wrap-up

Once projects were wrapped up, the teams presented their projects to the entire company. A team of internal judges gave out awards for “Best-In-Show” and “Most Immediately Impactful” and the entire company voted on a “Audience Choice” award.  After the final presentations, Scaliens gathered in a virtual stretch session to unwind and relax.

We hope this blog inspires you to host your own virtual hackathon or spark ideas for a hackathon that is in the works. If you’re interested in tackling new challenges and joining a collaborative and exciting team, take a look at all of our [open positions](https://scale.com/careers#jobs)!

Encouraging innovative thinking and making Scale AI better in practice with our first virtual Hackathon. 

Encouraging innovative thinking and tackling Scale AI's biggest challenges.

Scale Hackathon

Facundo Pepe

Back on September 16th, we introduced you to our <a href="https://scale.com/blog/introducing-our-new-api-docs" target="_blank">new API Docs</a>. Today, we’re pleased to follow up the original announcement with further details on how we implemented this with Next.JS, Tailwind CSS, and ReadMe. Furthermore, rather than simply describing how we used each platform and framework, we have packaged together an example for you to implement and build off of <a href="https://github.com/scaleapi/custom-documentation-example" target="_blank">in this repository</a>.
<h2 id="some-context%3A">Some Context:</h2>
At Scale AI, we are building the critical tools that make it easier to build and scale real-world AI systems. To provide our customers with the best experience possible on the Scale AI platform, we believe that a clear, simple, and memorable front-end, user interface, and user experience is critical. Scale’s design team has been hard at work building out a design system and improving customer-facing interfaces to bring front-end consistency across our platform.
Most important to this effort was to implement the design system across our different applications. We decided to use Tailwind CSS as a CSS framework and CSS Modules as a way to add custom elements when some customization was needed. To ensure our website and applications were highly performant and loaded quickly, we chose to use Next.js as our react framework.
<h2 id="why-tailwind%3F">Why Tailwind?</h2>
Scale’s design team have used Next.js and styled-jsx as a way of providing CSS styling since the initial launch of our website. While this was fine when there were fewer pages on the website, it proved to be hard to scale and maintain a design system without proper documentation. At this point the team faced two options: we could either spend a lot of time creating and improving our own CSS and writing documentation for it, or we could use a CSS framework that could adapt to our needs.
Since the Scale team prefers not to re-invent the wheel when it isn’t needed, we decided that using a CSS framework would yield the best results in the shortest period of time. Tailwind has allowed us to expedite the development of new layouts and pages while maintaining consistency across our website. Tailwind also has the advantage of having strong documentation behind it to support future development. Using Tailwind has also freed the team up to spend more time focusing on special details for each page like custom animations and interactions.
<h2 id="why-next.js%3F">Why Next.js?</h2>
Next.js is one of the best React Frameworks out today. Since the launch of <a href="https://nextjs.org/blog/next-9-3#next-gen-static-site-generation-ssg-support">SGG support</a> in March, it’s a no brainer for a fast and performant website and API docs.
Quoting them from their own <a href="https://nextjs.org/blog/next-9-3#next-gen-static-site-generation-ssg-support">blog post</a>:

Nextjs Quote

Next.js Quote

Next.js’s API routes also proved to be helpful as we used this to retrieve the search values made by the user. More on their API Routes [here](https://nextjs.org/docs/api-routes/introduction). 

## Why Readme?

As we highlighted in our previous [blog post](https://scale.com/blog/introducing-our-new-api-docs), we chose ReadMe as the platform for our API documentation because ReadMe allowed us to:

1. Unify the API reference and the Customer Dashboard documentation into a unified website.
2. Improve the developer experience. Instead of only showing static code examples, ReadMe enables developers to submit API requests directly with their Scale API key. ReadMe also has built in search functionality to make it easier for developers to find specific topics.
3. Democratize the content creation process. Instead of editing raw markdown, making a pull request, and then deploying to a server, we just had to update the text we needed in the ReadMe UI. Ultimately, the ReadMe platform provided a richer and more dynamic user experience.

## Sharing our example code with the developer community

We developed the new API documentation during Scale’s Hackathon. We are incredibly proud of what we accomplished in just one week and we wanted to share our solution with anyone who is looking to solve a similar problem. Rather than simply providing a step-by-step instruction for this solution, we have packaged together an example for you in our [Github repo](https://github.com/scaleapi/custom-documentation-example). 

We hope this sample code allows you to build your own beautiful and customized documentation.


How to customize your API Documentation’s front-end leveraging cutting-edge platforms and frameworks.

Customizing API Documentation with Next.JS, Tailwind CSS, and ReadMe.

Quick Start Guide to Customized API Documentation

Nishant Subramani

<h2 id="introduction">Introduction</h2>
The ML Team at Scale hosts a weekly reading groups where members choose papers from the broad AI/ML community and discuss them ranging from topics in Computer Vision to NLP to Active Learning. Here, we describe a brief summary of some of the insights we gained from various papers and how we aim to use some of that knowledge in future research projects and applications to Scale AI’s business.
<h3 id="discovering-useful-sentence-representations-from-large-pretrained-language-models"><a href="https://arxiv.org/abs/2008.09049" target="_blank">Discovering Useful Sentence Representations from Large Pretrained Language Models</a></h3>
Presenter: <a href="https://www.linkedin.com/in/nishantsubramani/" target="_blank">Nishant Subramani</a>
This is Scale’s first research paper and we wrote a <a href="https://scale.com/blog/discovering-useful-sentence-representations" target="_blank">blog post summarizing the paper</a>. This paper focuses on whether we can adapt pre-trained language models off-the-shelf as universal decoders. To be considered “universal,” a decoder must have an implicit representation for any target sentence such that it can recover that sentence exactly when conditioned on its representation. We investigated whether such representations exist and whether they can be easily discovered. Experiments show that not only do these representations exist for sentences from a variety of different genres, but also that our methods can recover these sentences almost perfectly without fine-tuning the underlying language model at all.

scale blog ml papers 1.jpg

Scale blog ml papers B_01

### [Scalable Nearest Neighbor Search for Optimal Transport](https://arxiv.org/abs/1910.04126)
__Presenter:__ [Anastasiia Alokhina](https://www.linkedin.com/in/anastasiia-alokhina-534b1014a/)

The authors of this paper look at the problem of nearest neighbor search with respect to Wasserstein distance. As Wasserstein distance has become popular in high-resource data environments and domains (images, text, etc.), search becomes prohibitively slow. As a result, approximation methods are important. The authors introduce a Quadtree variant called a Flowtree that formally shows better asymptotic accuracy and empirically show that it beats existing methods on a variety of different real-world datasets on either running time or accuracy.

Flowtree is a fast nearest neighbor search algorithm for Optimal Transport (Wasserstein-1 distance), which has a linear running time and speeds up SOTA performance by up to 7x. They do this by evaluating the optimal tree flow in the original metric space.

scale blog ml papers B_02

At Scale AI, we work with lots of high dimensional data such as documents and images for a variety of tasks. For these, we often need to find good representation spaces for these data points such that we can do fast and highly accurate similarity search in order to find outliers in datasets in order to find errors in labeling and find examples to label to improve our models using active learning. We think the flowtree method here could be invaluable in giving us the ability to use a different distance for this end and is something we will explore going forward. 

### [Structure-from-Motion Revisited](https://demuc.de/papers/schoenberger2016sfm.pdf)
__Presenter:__ [Hongbo Tian](https://www.linkedin.com/in/ppyht2/)

Structure-from-Motion(SfM) is a field of computer vision techniques which aim to reconstruct full 3-dimensional scenes from only a series of images. Accurate SfM algorithm can reproduce colored point clouds similar to LiDAR. The authors of this paper along with [Schönberger’s thesis](https://colmap.github.io/bibliography.html) introduced a robust computational pipeline for accurate dense 3D reconstruction. The authors also introduced [COLMAP](https://colmap.github.io/index.html) - an open-sourced CUDA accelerated software to reproduce the results.

scale blog ml papers B_03

Scale blog ml papers B_03

At Scale AI, our labelling pipeline is not subjected to the same online computational constraints compared to our customers. This allows us to pursue batch methods to enrich our 2D dataset, enabling our human labelers to capture 3D contexts from only images.

### [Representation Learning for Information Extraction from Form-like Documents ](https://storage.googleapis.com/pub-tools-public-publication-data/pdf/59f3bb33216eae711b36f3d8b3ee3cc67058803f.pdf)
__Presenter__: [Nishant Subramani](https://www.linkedin.com/in/nishantsubramani/)

The authors aim to extract structured information from form-like documents. They observe that forms have fields that often correspond to well-understood types and are often associated with a specific key phrase that bears a visual relationship with it. They also observe that key phrases in documents are drawn from a small vocabulary. Due to these observations, the authors build a system with two parts: candidate generation and scoring & assignment.

In the candidate generation phase each field type is associated with a candidate generator based on a cloud-based entity extraction service. Spans are then detected. In an invoice there are multiple dates. Every date in an invoice becomes a candidate for every date field in the target schema. This process is repeated for every target field.

In the scoring and assignment phase, the goal is to find the correct extraction candidate for each field. To do this, a score is computed for each candidate independently using a neural model. Next, the most likely scored candidate is assigned for each field. This process yields an independent learned representation based only on the candidates’ neighborhood (this is entirely independent of other candidates of fields). The neural scoring model is below.


scale blog ml papers B_04

Scale blog ml papers B_04

Our major takeaways are that structured extraction is not super well-studied by the academic community, but is relatively simple. The extraction system presented has promising accuracy and generalizes well to different domains. The representations have some interpretability as well. Ultimately some of the learnings from this paper could influence some of our work on [Scale Document](https://scale.com/document).

### [Big Transfer (BiT): General Visual Representation Learning](https://arxiv.org/abs/1912.11370)
__Presenter__: [Felix Lau](https://www.linkedin.com/in/felix-lau-0386641a/)

This paper focuses on how to transfer pre-trained representations to other tasks in the visual space. They propose a recipe called Big Transfer (BiT) and achieve very strong performance on ILSVRC-2012 (87.5% top-1 accuracy), CIFAR-10 (99.4%), and the Visual Task Adaptation Benchmark (VTAB; 76.3%).


scale blog ml papers B_05

Scale blog ml papers B_05

BiT has two phases: upstream and downstream. In upstream pre-training, the authors first investigate the scale of computational budget and how that affects performance. Secondly, they look at ground normalization and weight standardization, which helps significantly with small-batch and large-batch training alike. In downstream transfer, the authors propose a cheap fine-tuning methodology by using the BiT-HyperRule to select the most important hyperparameters for tuning as a function of the task’s intrinsic image resolution and number of datapoints. The hyperparameters they thought were important were training schedule length, resolution, and whether to use MixUp regularization.

There are a few takeaways from this method we found interesting. There’s a lot of value in balancing the computational budget and simplifying the experimental pipeline when needed. The BiT-HyperRule focusing on only a few hyperparameters was illuminating. We were interested in the dynamics of how large batches, group normalization, and weight standardization interplayed and were surprised at how poorly batch normalization performed relative to group normalization and weight standardization for large batches. The empirical strength of group normalization and weight standardization surprised us as well and we will start exploring these techniques in some of our computer vision models with lots of data available. However, the most impressive part was the empirical few-shot learning performance relative to the baseline. We are trying to build models that quickly adapt to new customers and domains and often struggle with fine-tuning, so these ideas are very top of mind.

### [Learning Loss for Active Learning](https://arxiv.org/abs/1905.03677)
__Presenter__: [Rishab Goyal](https://www.linkedin.com/in/rishgoyell/)

The authors present an active learning method that is a task-agnostic by attaching a loss prediction module to a target network and train it to be able to predict target losses of unlabeled inputs. The module can then provide insight into data that the model could predict incorrectly and thus improve the overall model by choosing just those examples to go and label.

scale blog ml papers B_06

Scale blog ml papers B_06

Their general process is that for every labeled example you predict the target and predicted loss for your prediction. Next, you take the unlabeled pool and pass them through the model to get predicted losses. Lastly you annotate the top-k data points and add them to the labeled training set. The loss prediction module uses a ranking loss.

As a data labeling company, we are very interested in active learning and how we can leverage our own strengths to better build ML models. We really liked the approach because it is widely applicable, but were concerned that there could be many degenerate cases. We weren’t convinced that the loss prediction module would be strong enough to offer significant improvements over uncertainty or coverage based methods such as entropy, query-by-committee, and coreset based methods.

### [LayoutLM: Pre-training of Text and Layout for Document Image Understanding](https://arxiv.org/pdf/1912.13318v5.pdf)
__Presenter__: [Malcolm Greaves](https://www.linkedin.com/in/malcolm-greaves/)

The authors try and solve the problem of automated understanding from scanned business documents. They use both computer vision and natural language processing techniques via textual and visual layout information for pre-training their system. Their system achieves SOTA on a set of layout analysis, receipt understanding, and document image classification tasks. They claim that their empirical performance is from their novel loss functions: a masked visual-language model (MVLM) loss and a multi-label document classification (MDC) loss.


scale blog ml papers B_07

Scale blog ml papers B_07

MVLM extends the masked language model loss from BERT by randomly choosing some words to mask. They then keep the masked token’s 2D positional embedding and forces the model to predict the word’s text. MVLM forces the model to understand language context and 2D relationships between words in the document. MDC assumes that documents have possibly several labels and is a standard cross-entropy loss. This is optional and seems to help on the document image classification tasks.

We found this paper interesting, but computationally demanding: alongside their loss functions, a key contribution of this paper is their pre-training routine. Without significant pre-training, these models do not seem to work well. This work requires off-the-shelf OCR and treats that as ground truth, which simplifies the work, but makes it less practical for document analysis practitioners. For the business document tasks that our ML team takes on, we often like to have greater control on the OCR task. It's sometimes advantageous to be able to adjust OCR to provide better downstream task-dependent results. We do a lot of document analysis at Scale, but can’t obfuscate or explain away the OCR component, which do produce some errors (e.g. with handwriting). Although this approach needs word-level annotations, Scale is in a unique position to get large amounts of these labor-intensive labels using our labeling platform.

### [Simplifying Models with Unlabeled Output Data](https://arxiv.org/abs/2006.16205v1)
__Presenter__: [Alexandre Matton](https://www.linkedin.com/in/alexandre-matton-73042b170/)

This paper focuses on tasks whose outputs have to obey some constraints. This happens in pseudo-code to code transcription for example. What we ultimately want is a code that compiles and runs without error. The usual way to address these tasks is to build a single end-to-end model and hope that it will learn the input-output mapping and understand the output constraints. The authors provide a new way to tackle these tasks by replacing the original model by two sub-models: a base predictor and a denoiser. The goal of the base predictor is to learn the mapping between the input and the output. The goal of the denoiser is to learn the constraints of the output space, in order to fix the base predictor’s output. 

scale blog ml papers B_08

Scale blog ml papers B_08

There is a reason why this decomposition works well. As the denoiser simplifies the job of the base predictor, it is now possible to increase the regularization of the base predictor without affecting the overall model performance. This in turn leads to solutions that generalize better on unseen data. The big win of this architecture is that the denoiser can be trained on unlabelled output data, which is often abundant (e.g. there is a huge amount of compiling code pieces freely available on github).

A big number of tasks actually impose constraints on the output (such as molecule generation, or translation from one language to another). At Scale, our ML team also tackles such problems. This paper provides a very general way to increase our downstream scores on them.

### Join our Team
If you’re interested in joining our growing engineering team, take a look at our [careers page](https://scale.com/careers) for open positions.


The machine learning research papers the Scale AI team read and discussed in Q3 2020.

Research Papers the Scale AI Team has Read in Q3 2020

Scale AI Machine Learning Digest - Q3 2020

Scale Team

At Scale AI, we envision a future where AI is as easy to build as any software. To help AI achieve its full potential, we are building the tools that make it easier to build and scale real-world AI systems. We began by focusing on the data - the foundation of all AI applications. With the Scale AI platform, customers can understand, visualize, and curate raw datasets, send raw data to be annotated, then debug model performance - all in one place.
From the beginning, we knew the Scale AI platform had to be API-based to enable teams working on AI and ML to automate their data pipelines and seamlessly integrate Scale into ML development cycles. Since our start in 2016, our engineering team has built incredible products to deliver large volumes of training data at high-quality, whether the data input is 3D sensor data, video, image, document, or text. Despite the continued evolution of our products and the robustness of our APIs, until recently, the API documentation itself was static, with limited interactions and not a lot of details to fully understand our products, set up effective workflows, and optimize for quality.
<h2 id="the-evolution-of-our-documentation%3A">The Evolution of our Documentation:</h2>
If you are a long-time Scale customer, you are probably familiar with this as our API documentation:

A screenshot of Scale's original API Documentation

Original API Documentation

Scale AI's Original API Documentation

The early documentation simply served as an API reference guide. The parameters were well spaced out, included relevant examples, and showed some example code using our SDKs. 

As we expanded the documentation to include content on key concepts, the customer dashboard, and other elements of the Scale AI platform, the documentation became increasingly clunky and awkward to navigate. The early documentation was also very static, with limited interactions, no search capability, and overall proved to be a less than ideal user experience. 

### Transition to ReadMe
Earlier this year, our Customer Experience and Field Engineering team identified [ReadMe](https://readme.com/) as an industry-leading documentation platform that could scale with our evolving needs. 

With ReadMe, we were able to realize three tangible benefits:

1. Unify the API reference and the Customer Dashboard documentation into a unified website.
2. Improve the developer experience. Instead of only showing static code examples, ReadMe enables developers to submit API requests directly with their Scale API key. ReadMe also has built in search functionality to make it easier for developers to find specific topics. 
3. Democratize the content creation process.  Instead of editing raw markdown, making a pull request, and then deploying to a server, we just had to update the text we needed in the ReadMe UI. Ultimately, the ReadMe platform provided a richer and more dynamic user experience.

The only limitation of the ReadMe platform out-of-the-box was the lack of front-end customization to ensure visual consistency between the documentation and the rest of our website. 

### Latest State:
The latest update to our documentation marries the content and functionality of ReadMe with our front-end to provide a better and more consistent experience between our API documentation and the rest of our platform. 

Newest State of API Documentation

Latest State of API Documentation

We accomplished this by effectively leveraging ReadMe as a content management system (CMS) on the back end. We then used [Tailwind CSS](https://tailwindcss.com/), a utility-first framework to build the UI for the front end, and [Next.JS](https://nextjs.org/)’s static generation feature to pre-render the page into HTML on the server ahead of each request. The HTML can be globally cached by a CDN and served instantly. Technical details on how we implemented this will follow in another blog post. 

## New Features:
### In-depth Customer Dashboard Documentation
As a Scale AI customer, the Dashboard is the central location to manage your account and your projects. The new in-depth Customer Dashboard documentation walks you through how to manage your account,  your projects, get an overview of submitted and completed tasks, and audit a random sample of tasks for quality.  

### Scale 101 Workflow Overviews
Every institution, company, and even team has their own ways of describing the data labeling workflow. To ensure every customer is successful on the Scale AI platform, we introduce the terminology we’ve developed in this section.  

Scale 101: Key Concepts

### Make API calls using your Scale API key directly from the Documentation
Instead of static code examples, this feature enables developers to submit live API requests directly using their Scale API key. 

API Key

### Search
As we continue to expand the documentation, we realize it can also become more challenging to navigate to and find details on a specific topic. To offer a faster and easier way to find keywords or key phrases, we have added a new search feature in our documentation powered by Algolia.    

Search Functionality

### Dark Mode
Dark mode reduces the light emitted by device screens while maintaining the minimum color contrast ratios required for optimal readability. To view the API documentation in dark mode, select “Dark” in your computer’s display settings or “Auto” to switch from light mode to dark mode at sunset or at a specified time. 

Light Mode/Dark Mode

## What this Means:
If you are an existing Scale customer, rest assured that the updates to the documentation do not impact any existing projects, batches, tasks, or workflows. If anything, we hope these updates will clarify the expected behavior, showcase our API’s flexibility, and enable easier task creation with more effective instructions to improve the overall quality of returned tasks. 

If you are just getting to know our products, we hope the updated documentation inspires you to come up with new project ideas. 

Introducing a new look and new features for our API Documentation

Introducing Scale AI's updated API Documentation

Introducing: Our New API Docs

Patrick Liu

At Scale AI, customers come first. This is a core credo that drives many of our daily decisions. We take the trust of our customers and the security of their data seriously, which is why we are pleased to announce the availability of our System and Organizational Controls (SOC) 2 Type 1 report.
As a cloud-based data annotation provider, our customers need to be confident that we securely handle their data. SOC reports are prepared by independent Certified Public Accountant (CPA) auditors based on the internationally recognized <a href="https://www.aicpa.org/content/dam/aicpa/interestareas/frc/assuranceadvisoryservices/downloadabledocuments/trust-services-criteria.pdf" target="_blank">Trust Services Criteria</a> framework developed by the American Institute of Certified Public Accountants (<a href="https://www.aicpa.org/" target="_blank">AICPA</a>). Achieving compliance with SOC 2 requires Scale demonstrating the achievement of their service commitments and system requirements based on the Security Trust Services Category. We take our commitments to security and the responsibilities they entail seriously, which is why we have partnered with CPA firm Schellman &amp; Company, LLC to provide ourselves and, most importantly, our customers with independent, third-party validation of our processes and controls to keep our systems and customer data safe.
Our commitment to enterprise-grade security is one reason leading ML teams such as NVIDIA, Toyota Research Institute, Open AI, and more rely on Scale AI to accelerate the development of their AI applications. Scale continues to invest in security best practices, and to that end, we are pursuing a Type 2 report moving forward. If you’d like further details on our SOC 2 report, please contact us at <a href="mailto:security@scale.com" target="_blank">security@scale.com</a> or <a href="mailto:sales@scale.com" target="_blank">sales@scale.com</a>.

At Scale AI, we take the trust of our customers and the security of their data seriously. We are pleased to announce the availability of our SOC 2 report. 

Reaffirming Our Commitment to Building Customer Trust: Scale+ SOC 2

Reaffirming Our Commitment to Building Customer Trust: Scale + SOC 2

Alexandr Wang

When we founded Scale in 2016, we envisioned a future where AI was as easy to build as any other software. Four years on, we’ve seen incredible success bringing that vision to life for a huge range of businesses across self-driving, natural language, ecommerce, robotics, and many other fields. As we continue to develop our ambitious roadmap to accelerate AI development, I’m delighted to welcome Brad Porter—formerly Amazon’s VP and Distinguished Engineer of Robotics—as our first CTO.
<h3 id="our-vision-for-scale">Our vision for Scale</h3>
I got hooked on machine learning from a young age. As a student, I’d build models in my free time that predicted how hard it would be to find parking outside my house, or whether I needed to restock my fridge on the way home.
The beauty of AI systems is their flexibility—define a goal, and they learn an optimal process themselves. But I soon realized that AI is much harder to deploy than it should be. It was incredibly cumbersome to find enough data to train my models. And when I moved, my parking model needed retraining—so I had to find more data all over again.
Other fields of software, like app development, have an entire ecosystem of support tools that streamline development, abstracting away complexity and automating basic workflows.
You don’t need a PhD to build a website. What if the same was true for AI?
<h3 id="building-the-technology-stack-for-ai-development">Building the technology stack for AI development</h3>
Brad and I first met a year ago. I had made it a point to get to know all of the people out there who were doing production AI in meaningful ways. Brad, leading robotics at Amazon, was at the top of the list.
Brad is a very humble guy. It is striking when you meet him. Many technology leaders have a bit of show and flair to them, but Brad immediately reminded me of the kind of leader that I strive to be: exceedingly humble, extremely technical, excited by ideas (both good and bad), and insatiably curious to understand how things work. As we were getting to know each other, he told me that he’s been in love with things that can fly since he was a kid: paper airplanes, kites, boomerangs, radio-controlled planes, flying soda cans, and more. As a kid who was personally obsessed with math and physics out of an intrinsic curiosity for the world, I realized Brad and I had a lot in common.
When we were talking about successful teams, I quickly realized I had a lot to learn from Brad. One core tenet of our work at Scale is our belief that ambition shapes reality. It is something we try to live every day. Unprompted, in one of my email correspondences with Brad, he wrote: “One thing all startups need is ambition. In my experience, ambition is the greatest limiter.”
Brad has a reputation as an incredible leader. One indication is former teammates who are fiercely loyal and incredibly effusive. Every person I have spoken to who has worked with Brad, either at Netscape, Tellme, or Amazon, could not stop raving about him. The same themes continued to pop up—humble, compassionate, dedicated to his teams, curious, and deeply technical.
Then, there’s all of Brad’s technical accomplishments! More than almost anyone, Brad knows what it takes to productionize AI systems and deploy them at global scale in complex, changing environments. With his experience leading robotics at Amazon, he managed one of the most impressive uses of machine learning at scale—the over 200,000 robots powering Amazon’s unrivaled global warehouse and distribution infrastructure. He also has deep expertise in natural language, working with AI and language scientists at Tellme Networks to accelerate the rollout of speech recognition, processing and dialog management technologies. I’m thrilled he has decided to join us.
Brad and I both agree that getting effective, accurate and unbiased data is currently the hardest problem in machine learning. At Scale, we’ve built tools to eliminate those frictions. We’ve built an API that integrates seamlessly with our customers’ workflows, and our own ML models automate much of the annotation process and help ensure we deliver ground-truth data of the highest quality.
As our CTO, Brad will help continue to improve the efficiency and sophistication of our technologies and processes by orders of magnitude, building new tools to continue to accelerate the development of AI systems. He will manage our entire technology stack, with combined ownership of product direction and engineering, to turbocharge our ambitious technology roadmap.
Brad understands first hand the need for better tools that will become the backbone of AI development. Under his leadership, we’ll be able to accelerate our development of tools that can support the AI development lifecycle—building the full development stack for AI.
<h3 id="scale%E2%80%99s-network-effects-and-future-of-ai">Scale’s network effects and future of AI</h3>
Right now, AI is at a crossroads. With systems like AlexNet, AlphaGo and GPT-3, AI research has made incredible strides. But demand for data annotation is far outstripping the supply of skilled labelers. And the data to be annotated is only getting more complex as AI models become more sophisticated.
With data bottlenecks persisting, the delta between what AI could do in theory and in practice is only growing. The field desperately needs better technology to overcome these barriers.
But a new era of AI has begun. AI is about to enter a wave of productionization. The industrialization of AI—with global spending on AI systems <a href="https://www.businesswire.com/news/home/20180919005045/en/Worldwide-Spending-Cognitive-Artificial-Intelligence-Systems-Forecast" target="_blank">predicted</a> to triple between 2018-2022—is beginning to take shape.
What’s holding it back is a lack of infrastructure. I believe Scale will play a major role shaping that future. Thanks to our own machine learning technologies, our data annotation products become more sophisticated with every annotation instance they encounter. That provides powerful network effects that benefit all our customers. As we grow, the impact on the deployment of AI will be transformative.
I look forward to working with Brad as he helps drive our progress on that journey.
If you’d like to join Brad’s engineering team, you can find all our open roles on our <a href="https://scale.com/careers" target="_blank">careers page</a>.

Brad Porter—formerly Amazon’s VP and Distinguished Engineer of Robotics—joins Scale AI as our first CTO

Scale AI welcomes Brad Porter as our first CTO

Welcoming Brad Porter to Scale AI as our first CTO

Nivedita Suresh

<h2 id="ai%2Fml-research-at-scale-ai%3A">AI/ML Research at Scale AI:</h2>
Scale AI is dedicated to accelerating the development of AI applications. As part of this mission, Scale AI’s growing AI/ML Team aims to meaningfully contribute to the AI ecosystem with research and resources that can bring significant and practical benefits to our customers and the broader research community.
Research and experimentation happens throughout Scale. In some cases, our contribution is in areas where the bar has already been set high. Some examples include the open-source datasets we have released with various partners such as <a href="https://pandaset.org/" target="_blank">PandaSet</a>, <a href="https://www.nuscenes.org/" target="_blank">nuScenes</a>, and <a href="https://scale.com/resources/download/nlp-dataset-balanced-conll" target="_blank">CoNLL Balanced</a>. As our AI/ML team continues to grow, we aim to contribute in entirely new spaces by publishing cutting edge research papers. In this blog post, we spotlight a research paper we recently submitted.
The research paper was a collaboration with <a href="https://www.linkedin.com/in/nivedita-suresh/" target="_blank">Nivedita Suresh</a> of Arrive. <a href="https://arriveorigin.com/" target="_blank">Arrive</a> links multi-omics and AI cell characterization to discover hidden patterns in data and improve outcomes in precision medicine.
<h2 id="conditioning-unconditional-language-models-to-recover-sentences-without-fine-tuning%3A">Conditioning Unconditional Language Models to Recover Sentences without Fine-Tuning:</h2>
Pretrained language models are used as encoders with widespread success for a wide variety of NLP tasks. These models learn useful representations that can be directly applied to certain tasks. Despite the success of these models on the encoder, there isn’t much research out today on whether these large pretrained language models can be used as universal decoders. To be considered “universal,” a decoder must have an implicit representation for any sentence, such that it can recover that sentence exactly when conditioned on its representation.
Task-specific decoders are typically trained from scratch for each task. In this research, we propose a method to try and learn useful representations of sentences using large, pretrained, transformer-based decoders. Our work investigates whether a universal decoder is even possible. If we had access to such a decoder, we could use it for a wide variety of sequence generation tasks without having to retrain or rebuild the decoder on new data. This could open up a new paradigm for low-resource machine translation, summarization, image captioning, and dialog generation.
<h2 id="the-sentence-space%3A-how-language-models-represent-sentences">The Sentence Space: How Language Models Represent Sentences</h2>
Transformer-based language models like ELMo, BERT, and GPT-2 have replaced word vectors as off-the-shelf general-purpose encoders for natural language understanding. These models are trained on large amounts of text data and represent a sentence as a sequence of hidden states that come from a final layer of the transformer model. Representations in this sentence space are sequence length dependent, making comparisons between sentences with differing lengths inequitable and measuring the efficacy of using an unconditional language model as a universal decoder impossible.
To resolve these issues and to make analysis easier, we propose to reparametrize the original sentence space into a lower-dimensional and sentence length agnostic vector space. We do this by adding a bias term to the fixed language model and finding the representation that minimizes the cross entropy loss of the sentence. This reparameterization now gives us the ability to project sentence tokens to representations (sentence encoding) and recover sentences from the representation (sentence recovery) via the fixed language model. We develop three representation injection mechanisms and inject biases at three different locations.

Figure 1

Sentence space representations of transformer-based language models.

## Experiments:
Since we are testing the efficacy of a large pretrained language model as a universal decoder, we measure how well a representation can recover a sentence when injected into a language model. We measure performance across sentences from 4 different genres: books, news articles, Wikipedia, and movie dialogs. We conduct controlled experiments, varying the representation injection mechanism, representation injection location, initialization, and dimensionality of the representation.

## Results:
We find that our representations recover sentences nearly perfectly even across genres -- achieving a BLEU score of over 98.


BLEU performance on LRC stratified by genre for different dimensionalities of z.

Figure 5: BLEU performance

BLEU Scores

In addition, we observe that the intrinsic dimension of the representation space is the model’s latent dimension, which indicates that the language model uses its entire capacity to represent sentences.

Plot of sentence length vs. BLEU score on LRC for experiment IV with error regions.

BLEU

Sentence length v. BLEU score

Additionally, our interpolation experiments reveal that the representation space has some human-understandable meaning. Our learned representations seem to have some synonym awareness. In the first sentence pair example below, the word “tale” transforms to the word “story” and the word “long” transforms to “long-running” when referring to a war. In the second example, we observe some syntactic awareness at the 0.7 mixture level. The syntax of the first sentence is retained with mostly words from the second sentence. 

Interpolation

Linear interpolations between perfectly recovered pairs.

## Broader Impact:
This work informs us that we can discover meaningful, sentence-length agnostic sentence representations, hinting at the possibility of a “universal” decoder. Such a decoder would improve low-resource sequence generation task performance and allow for considerable parameter sharing in memory and data-limited environments.

## Conclusion:
For further details on how we set up the experiment, results, and analysis, you can find the full paper on [arXiv](https://arxiv.org/abs/2008.09049). If you’re interested in joining our growing machine learning team, you can find all of the open positions on our [careers page](https://scale.com/careers?location=All&team=Engineering). 

Learning useful sentence representations from large transformer-based language models.

Discovering Useful Sentence Representations from Large Pre-trained Language Models

Discovering Useful Sentence Representations from Large Pre-Trained Language Models

James Lennon

<h2 id="introduction">Introduction</h2>
Machine learning is a once-in-a-generation technological shift that is creating huge value by unlocking insights from data. But algorithmic bias remains a major concern as ML becomes more widespread. Unless ML models are trained on representative data, they can develop serious biases, significantly harming underrepresented groups and leading to ineffective products. We investigated the CoNLL-2003 dataset—a standard for building algorithms that recognize named entities in text—and found that the data is highly skewed toward male names. Using Scale’s technology we were able to systematically mitigate this bias by:
<ol>
<li>Enriching the data to illuminate the hidden biases</li>
<li>Enhancing the dataset with underrepresented examples to mitigate gender bias</li>
</ol>
A model trained on our enhanced CoNLL-2003 dataset has both less bias and better performance—demonstrating how bias can be removed without any changes to the model. We’ve open-sourced the Named Entity Recognition annotations for the original CoNLL-2003 dataset and our augmentation of it <a href="https://scale.com/resources/download/nlp-dataset-balanced-conll" target="_blank">here</a>.
<h2 id="algorithmic-bias%3A-ai%E2%80%99s-weakness">Algorithmic bias: AI’s weakness</h2>
Today, thousands of engineers and researchers are building systems that teach themselves how to achieve significant breakthroughs—improving road safety with self-driving cars, curing diseases with AI-optimized treatments, and combating climate change by managing energy usage.
But the power of self-learning systems is also their weakness. Because data is the foundation of all machine learning applications, learning from imperfect data can lead to biased outputs.
The power of AI systems means their potential to cause harm if they are biased is significant. Recent protests against the police brutality that has led to the tragic deaths of George Floyd, Breonna Taylor, Philando Castile, Sandra Bland and many others are an essential reminder of the systemic inequalities in our society that AI systems must not perpetuate. But, as we know from countless examples—whether it’s image search results <a href="http://www.mjskay.com/papers/chi_2015_gender-bias-in-image-search.pdf" target="_blank">perpetuating gender stereotypes</a>, offender management systems <a href="https://www.propublica.org/article/machine-bias-risk-assessments-in-criminal-sentencing" target="_blank">discriminating against black defendants</a>, or facial recognition systems <a href="https://www.nist.gov/news-events/news/2019/12/nist-study-evaluates-effects-race-age-sex-face-recognition-software" target="_blank">misidentifying people of color</a> – we have a long way to go before we can say the problem of AI bias has been solved.
Bias is prevalent because it is so easily introduced. It creeps up, for example, in gold-standard open-sourced models and datasets that are the foundation of huge volumes of work in ML. The word2vec sentiment dataset for building other language models is <a href="https://blog.conceptnet.io/posts/2017/how-to-make-a-racist-ai-without-really-trying/" target="_blank">skewed by ethnicity</a>, and word embeddings—the way an ML algorithm represents words and their meanings—have been found to contain <a href="https://arxiv.org/pdf/1608.07187.pdf" target="_blank">strongly biased assumptions</a> about what occupations women are associated with.
The problem—and at least part of the solution—lie in data. To illustrate this, we’ve conducted an experiment into one of the most popular datasets for building systems that can recognize named entities in text: CoNLL-2003.
<h2 id="what-is-named-entity-recognition%3F">What is Named Entity Recognition?</h2>
Named-Entity Recognition (NER) is one of the fundamental building blocks of natural language models—without it, online searches, information extraction, and sentiment analysis would be impossible.
At Scale AI, our mission is to accelerate the development of AI. Natural language is one of our main areas of focus. Our Scale Text offering includes NER—which involves annotating text according to a predefined list of labels. In practice, this might help major retailers analyze how their products are being discussed online, among other applications.
Many NER systems are trained and <a href="https://paperswithcode.com/sota/named-entity-recognition-ner-on-conll-2003" target="_blank">benchmarked</a> on CoNLL-2003—a dataset of roughly 20,000 sentences from Reuters news articles that are annotated with attributes such as ‘PERSON’, ‘LOCATION’, and ‘ORGANIZATION’.
We wanted to explore if the data was biased. To do this, we used our Scale AI labeling pipeline to categorize all the names in the dataset, asking whether they could be male, female, or either, inferring gender based on the traditional use of the name.
We found a dramatic difference. Male names are mentioned almost five times more than female names, and less than 2% of names were gender-neutral:
<figure class="expand-image">
 <img src="//images.ctfassets.net/1qgw036lpkfx/23DTaOdXIuKthLUAEG4Pdg/335f2763d6373c91cf113a60bd100282/chart-2.svg" />
</figure>
This is because news articles, for societal reasons, tend to contain mostly male names. But an NER model trained on this data will be better at picking out male names than female. For example, search engines use NER models to help classify names in search queries to give more accurate results. But deploy a biased NER model, and the search engine will be worse at identifying names of women than of men—exactly the kind of subtle, pervasive bias that can creep into many real-world systems.
<h2 id="a-novel-experiment-to-reduce-bias">A novel experiment to reduce bias</h2>
To illustrate this, we trained an NER model to investigate how this gender bias would affect its performance. We built a name extraction algorithm to pick out PERSON labels using <a href="https://spacy.io/" target="_blank">spaCy</a>, a popular NLP library, and trained the model on a subset of the CoNLL data. When we tested the model on novel names in the test data that weren’t present in the training data, we found that the model was 5% more likely to miss a new female name than a new male name—a significant discrepancy in performance:
<figure class="expand-image">
 <img src="//images.ctfassets.net/1qgw036lpkfx/2ZYmbgILyiKAI4tjSgIDph/9d431f098a830c0c3c0df5b820ab0b1b/chart-3.svg" />
</figure>
We saw similar results when we used our model on the template “NAME is a person”, substituting the 100 most popular male and female names for each year of the US census—the model performs significantly worse on female names for all years of the census:
<figure class="expand-image">
 <img src="//images.ctfassets.net/1qgw036lpkfx/lsDqk1QgcCnL2vt1jxbdO/fce68bdac2e729f28e97392e8ca31a8e/chart-4.svg" />
</figure>
Crucially, biased training data means errors are skewed toward underrepresented categories. This census experiment illustrates this a second way, too: the model’s performance degrades significantly after 1997—the cut-off point of the Reuters articles in the CoNLL dataset—because the dataset is not representative of the popularity of names in the years since.
Models learn to fit the trends of the data they are trained on. They can’t be expected to perform well on cases for which they have seen few examples.
<h2 id="if-you%E2%80%99re-de-biasing-the-model%2C-it%E2%80%99s-too-late">If you’re de-biasing the model, it’s too late</h2>
So how do we fix this?
One way is to try and de-bias the model—perhaps by post-processing the model, or adding an objective function to mitigate bias, leaving it to our model to figure out the details.
But there are several reasons why this is not the best approach:
<ol>
<li>Fairness is a highly complex issue that we can’t expect an algorithm to define on its own. Research has shown that training an algorithm to perform equally on all subsets of the population will not ensure fairness and will <a href="http://blog.mrtz.org/2016/09/06/approaching-fairness.html" target="_blank">cripple the model’s learning</a>.</li>
<li>Adding extra objective functions can hurt model accuracy, causing a tradeoff. Better instead to keep the algorithm simpler and ensure the data is balanced—improving model performance and avoiding the tradeoff.</li>
<li>It’s unreasonable to expect a model to perform well on cases for which it’s seen few examples. The best way to ensure good results is to improve the diversity of the data.</li>
<li>Attempting to de-bias a model with engineering techniques is expensive and time-consuming. It’s much cheaper and easier to train your models on unbiased data in the first place, freeing up your engineers to focus on applications.</li>
</ol>
Data is only one part of the bias problem. But it is foundational, affecting everything that comes after it. That’s why we think it holds the key to some of the solution, providing potential systematic fixes at the source. Unless you explicitly label for protected classes, like gender or ethnicity, it is impossible to properly mitigate these classes as a source of bias.
This is counterintuitive. Surely, if you want to build a model that doesn’t depend on sensitive characteristics, like gender, age or ethnicity, it’s best to omit those properties from the training data so the model can’t take them into account?
“Fairness by ignorance” actually makes the problem worse. ML models excel at drawing inferences across features—they don’t stop doing this just because we haven’t explicitly labeled those features. The biases simply remain undetected, making them harder to remove.
The only robust way to deal with the problem is to label more data to balance out the distribution of names. We used a separate ML model to identify sentences in the <a href="http://kdd.ics.uci.edu/databases/reuters21578/reuters21578.html" target="_blank">Reuters</a> and <a href="https://archive.org/details/BrownCorpus" target="_blank">Brown</a> corpora likely to contain female names, and then labeled those sentences via our NER pipeline to augment CoNLL.
The resulting dataset, which we’ve called CoNLL-Balanced, has over 400 more female names. When we retrained our NER model on it, we found that our algorithm is no longer biased to perform worse on female names:
<figure class="expand-image">
 <img src="//images.ctfassets.net/1qgw036lpkfx/4vOWZTzsDU7MxxK870EiE1/e685239cfbc507405d64072f466ebbee/chart-9.svg" />
</figure>
Not only that, but the model also performed better at identifying male names, as well.
This is an impressive demonstration of the power of data. Mitigating bias at the source meant we did not have to make any adjustments to our ML model—saving engineering time. And we achieved this without any tradeoff in our model’s performance—in fact, its performance slightly improved.
<figure class="expand-image">
 <img src="//images.ctfassets.net/1qgw036lpkfx/3vc17brFOvBpwmPxJvThR/d74fe758f2a81a30db5458ba8bb7bd4d/image__14_.png" />
</figure>
To allow the community to build on our work and mitigate gender bias in models built on CoNLL-2003, we’ve open-sourced Scale AI’s augmented dataset, which also includes gender information on <a href="https://scale.com/resources/download/nlp-dataset-balanced-conll" target="_blank">our website</a>.
The AI/ML community has its own diversity problems, but we are cautiously excited by these results. It suggests how we might be able to offer a technical solution to a pressing social problem—provided we tackle the issue head on, illuminating the hidden biases and improving the model’s performance for everyone.
We’re now exploring how we might apply this approach to another highly sensitive attribute—ethnicity—to explore whether we can create a robust framework for eliminating dataset bias that scales to other protected classes.
It also illustrates why we at Scale AI pay so much attention to data quality. Unless data is provably accurate, balanced, and unbiased, there’s no guarantee that models built on it will be safe and accurate. And without that, we won’t be able to build the transformational AI technologies that benefit everyone. If you are developing AI technologies and would like help in balancing your own datasets to ensure your models work for everyone, <a href="https://scale.com/sales" target="_blank">reach out to us</a>.
Acknowledgements:
The CoNLL 2003 Dataset referenced in this blog post is the Reuters-21578, Distribution 1.0 test collection, available on the project page for the original 2003 experiment: <a href="https://www.clips.uantwerpen.be/conll2003/ner" target="_blank">https://www.clips.uantwerpen.be/conll2003/ner/</a>.

Scale AI's platform can help mitigate bias before training a model.

The importance of detecting and mitigating bias from training data before even starting to build an algorithm.

If You’re De-Biasing The Model, It’s Too Late

We founded Scale to create the infrastructure needed to build AI in any industry, by anyone. We started tackling this complex problem at the root — turning raw data into high quality training data for models. In this pursuit, we spent the last four years building ML-augmented annotation products for all data types, expanding our solutions for major industries, and making significant technological strides in <a href="https://scale.com/blog/how-we-scale-machine-learning" target="_blank">scaling our use of ML</a> in doing so.
But the problem of building effective, accurate, unbiased ML models still remains. To do this, aggregate metrics in ML are not good enough. Better ML starts with understanding your data in depth. To improve production ML, you need to understand their qualitative failure modes, fix them by gathering the right data, and curate diverse scenarios.
<ol>
<li>Before training a model, ML engineers must curate and sample their data, ensuring that they have the right data to solve a specific problem. This process is too often very manual. For example, to teach a self-driving car how to handle left turns, an ML team has to manually crawl through their driving sequences to isolate examples of left turns to create a training dataset. The data also needs to be representative of the ground truth of the problem you’re trying to solve. If you’re building a model to assign a gender to faces, you have to ensure representative data for all genders to have unbiased outputs. Again, that’s too often a manual and highly inefficient process.</li>
<li>After training, ML teams test and benchmark model performance, ensuring that the test dataset is also sufficiently representative of the problem they are trying to assess its performance against. For example, a model learning how to tell cars apart from pedestrians needs enough of both examples for accurate benchmarking. This requires ML engineers to spend significant amounts of time building one-off UIs to chart and share performance data.</li>
<li>After deployment, debugging the model, identifying failure modes and fixing them. Too often issues in data only arise after a model has entered deployment – requiring time-consuming debugging. One Scale customer, for example, found that their vehicle recognition algorithm didn’t perform well in certain environments – it turned out that the model was trained on a dataset where vehicles were mostly in the bottom of the image, so the model associated “bottom of the image” with “likelihood of being a car.”</li>
</ol>
The Scale team has been working to productize the concept that Andrej Karpathy calls “<a href="https://www.youtube.com/watch?v=oBklltKXtDE&amp;feature=youtu.be&amp;t=475s" target="_blank">Operation Vacation</a>.” Nucleus is a new way—the right way—to develop ML models, helping us move away from the concept of one dataset and towards a paradigm of collections of scenarios and giving ML engineers the ability to automate time-consuming manual steps in the ML development process. 
<img src="//images.ctfassets.net/1qgw036lpkfx/01d1iYza6qIJQco0C6bIi0/e4f6510aa29f322f975b2c4cfc0c2dc1/dev_cycle_diagram_03.svg" alt="ML Development Cycle">

[Scale Nucleus](https://scale.com/nucleus) provides advanced tooling for understanding, visualizing, curating, and collaborating on your data – allowing teams to build better ML models via a powerful interface and APIs. With Scale Nucleus, you can:
- __Visualize your dataset__, ground truth, and model predictions to improve model performance
- __Curate interesting slices within your dataset__ for active learning and identifying key edge cases
- __Upload and choose data to be annotated__ for rare event mining and dataset balancing
- __Search your data__ based on metadata or ML-produced attributes
- __Identify edge cases__ through visual search
- __Measure key metrics__ like dataset balance, class correlation, and confusion, via a powerful insights tab that shows the overall health of the data
- __Debug__ model performance
- __Share your dataset__ seamlessly to provide a single source of truth for data within your team.

pandaset - query

Scale Nucleus Query

Scale Nucleus is directly integrated with Scale AI’s data labeling pipeline – allowing teams to fix any issues in their data at the source. Nucleus currently supports Image data, with support for 3D Sensor Fusion, Video, Text and Document data coming soon.

We built Scale Nucleus to allow data scientists and ML engineers to manage data more efficiently and increase the marginal value of their data. Deeper insight into a dataset’s features can have several transformational effects on the development of AI:

- For engineers, the easier visualization and curation of datasets lowers the barriers to entry to building ML systems.
- Spotting hidden modes of failure in datasets before deployment (such as a set of driving sequences that doesn’t include any sequences at night) makes it much easier to train high-quality models, and provides a robust way of eliminating issues like bias at source.
- The ease of debugging could significantly improve iteration speed on model fine-tuning post-deployment. 

We are excited to take the next step in building the infrastructure to enable efficient, accurate, and unbiased ML development. If you’d like to join us in this journey and try Scale Nucleus, contact us at nucleus@scale.com or sign up on our [website](https://scale.com/nucleus#sales). 


Scale Nucleus: The mission control for your data

Introducing Scale AI's latest product: Scale Nucleus

Introducing: Scale Nucleus

June Lee

Fostering connection when everyone is remote can be tricky. Wind back to four months ago-- it was easy to high five a coworker, give surprise birthday cakes outside of meeting rooms, and hop down to the juice shop downstairs for a walk and talk. Shelter-in-place restrictions pose new challenges that inspired our people team to think of creative ways to keep the team connected.
Our initiatives centered around the following dimensions:
<ul>
<li>Social Connection - How do we facilitate interaction between people across different teams and departments?</li>
<li>Wellness - How do we encourage people to take care of themselves and remain mentally well during uncertain times?</li>
<li>Alignment - How can we keep teams aligned, engaged, and focused on company-wide goals and objectives?</li>
</ul>

The People Team's initiatives and how they map to social connection, wellness, and alignment.

People Team Initiatives

For the purpose of this blog post, we chose to highlight and share our social connection initiatives. We are happy to share these ideas with other companies looking to provide employee social support.

## Wellness Programs

The pandemic threw a curveball of new work environments, screen-focused social interactions, along with uncertainty around the current economy. To help our employees prioritize their health and well-being, we launched a wellness program. We created a space for our employees to participate together, foster camaraderie, and not feel alone in their personal wellness endeavors. 

### April Wellness Program: Building Healthy Habits

An example of our April Wellness Tracker.

April Wellness Tracker

__Details of the Program__ 
The program reminded employees to track key health criterias to build healthy habits. In a two-week span, participants recorded if they met the following criterias:
<ul>
 <li>7+ hours of sleep</li>
 <li>Drink 64 oz + of water</li>
 <li>20+ minutes of exercise</li>
 <li>10+ minutes of mindfulness</li>
</ul>

__Social Interaction__ 
We encouraged social interaction through a leaderboard, seeing how we stack rank against each other. We also had a slack channel dedicated to this challenge to share knowledge with each other and asking, “Does tea count towards our water intake?” We could also commiserate on the channel about how drinking a lot of water was difficult.

We ended the challenge with a virtual celebration highlighting winners for:
<ol>
 <li>Highest consecutive points</li>
 <li> Category leaders </li>
 <li> Week 2 leaders (to motivate late joiners) while celebrating everyone who participated</li>
 </ol>
Overall, the participation rate was 20 percent of the whole company.

[Gabby Zetino](https://www.linkedin.com/in/gabbyzetino/), our sales development representative, shares her experience with the challenge:

Quote from Sales Development Representative, Gabby Zetino, on participating in the April Wellness Challenge.

Gabby Zetino Quote

### May Wellness Program: Bingo Board and Scale Sponsored Events


May Wellness Program

May Wellness Challenge Tracker

__Details of the Program__
 
This month, in honor of Mental Health Awareness Month, we assembled a mental health bingo board which highlighted a list of activities employees can do at their own pace. If an employee completed ten of the activities, they were entered into a raffle. 

[Taisia Karaseva](https://www.linkedin.com/in/taisiakaraseva/), our Pilots & POC lead, shares how the bingo board pushed her to reflect on wellness habits along with the Scale-sponsored event:

Quote from Pilots & Proof of Concept Lead, Taisia Karaseva, on participating in the May Wellness Challenge.

Taisia Karaseva Quote

__Social Interaction__ 
We had a column dedicated to social interactions: “Stay Connected.” Also, Scale sponsored events on the bingo board’s first row: 

- Stress Relief Workshop + Guided Meditation - we brought in a self-care coach and meditation instructor for a 15-minute workshop on how to relieve stress and 15-minute meditation to regain balance.
- Yoga with Scale - we brought in a yoga instructor for a quick 30-minute session in the morning
- Scale’s Creative Time- we set aside time to hang out and work on something creative together (paint, write, dance, bake etc)

## Scale Themed Trivia Nights 
Scale Trivia Nights were born out of shelter-in-place and is one of our highly attended events hosted by our workplace operations team, [Monica Breton](https://www.linkedin.com/in/monicabreton1/) and [Eileen Ho](https://www.linkedin.com/in/eileen-h-66956b121/).

<figure class="expand-image">
 <img src="//images.ctfassets.net/1qgw036lpkfx/4Md0DMBzQBhbhrUAJCTuuy/af96311d89acc34c497efd0eed5736b6/05.jpg" />
</figure>

__Details of the Program__
 
The program administrator prepared questions in advance, selected employees into groups randomly and sent them off to their breakout virtual rooms. Similar to regular trivia, each team had a lead and a team name. The trivia administrator gave category questions to the leaders via slack and the teams in their [breakout rooms](https://support.zoom.us/hc/en-us/articles/206476313-Managing-breakout-rooms).

Then within the breakout rooms, teams discussed and laughed about the answers they’ll give and relayed the answers back to the program administrator. 

__Social Connection__
 
[Hannah Moreno](https://www.linkedin.com/in/hannahmoreno1/), our marketing operations specialist, shares a few words on her experience with Trivia Night:

Quote from Marketing Operations Specialist, Hannah Moreno, on participating in Scale Trivia.

Hannah Moreno Quote

## Speed Scaling
To play on the words of “speed dating”, at Scale, [Margo Wheeler](https://www.linkedin.com/in/margowheeler/), our people operations manager, hosted a networking event, “Speed Scaling” for us to connect with employees across different teams with fun icebreaker questions.

<figure class="expand-image">
 <img src="//images.ctfassets.net/1qgw036lpkfx/68LqkO7xM6vHTy0khHIvTQ/5363200049df2be41dcc482c267e243c/06_B.png" />
</figure>

__Details of the Program__
 
To do this remotely, Margo broke us into different breakout groups and posted the questions to us in real time. We popcorned around in a circle answering questions. Our answers organically led us to unprompted conversations. When we had all spoken and experienced just the right amount of awkward silence, we moved onto the next question to continue learning about each other. 

For quarantine edition, we also included these questions: 
- What are you most looking forward to doing once Shelter-In-Place is lifted? 
- What’s the best meal you’ve cooked (or someone’s cooked for you) during Shelter-In-Place? … or baked if you’re a baker 
- What is your Go-To Quarantine Snack?

__Social Connection__
 
[Guido Maliandi](https://www.linkedin.com/in/gmaliandi/), our engineering manager, had attended both the live and the remote Speed Scaling sessions. He shares his experience with both:

Quote from Engineering Manager, Guido Maliandi, on participating in the Speed Scaling initiative. 

Guido Maliandi Quote

## Small Groups Scale Socializing

Socializing remotely could be difficult for a lot of people, especially in larger groups. To foster a genuine connection, smaller groups of one-on-one or groups with no more than four were scheduled by Eileen Ho, our program manager for Scale Socializing. 

![Coffee Chat](//images.ctfassets.net/1qgw036lpkfx/5iM0REWMyUdUBEex58bXtR/a2ee98859f029f909c9afbd32c953000/08_B.png)

__Details of the Program__
 
We sent a survey out to the company so that anyone who chooses to can opt in to socialize. There are logistical nuances we consider such as: 
- Who they want to meet (new hires, cross functional team members) 
- How often they want to socialize 
- If they prefer one-on-one settings or groups of three. 
In addition to [donut](https://www.donut.com/) integration we’re already using, this initiative facilitates further connections for employees. 

__Social Connection__
 
Our field engineering manager, [Shaun VanWeelden](https://www.linkedin.com/in/svanweelden/) has participated in several of these virtual coffee chats. Shaun shares his experience with Scale Socializing below:

Quote from Field Engineering Manager, Shaun VanWeelden, on participating in Scale Socializing.

Shaun VanWeelden Quote

## Slack Channels and Prompts: Scalien Share

Slack is one of the main sources we use to keep in touch, so it only made sense to take advantage of it when we started “Scalien Share.” Similar to “Show and Tell” it is a fast way to take a step away from work and open up about ourselves. In addition to “Scalien Share”, we keep track of our interest-group slack channels for new hires to join. Most of these groups are grassroots-led such as #thebookclub, #cats, and #scaleai-women.
![Scaliens Share: Recent Foods](//images.ctfassets.net/1qgw036lpkfx/hGjy9wMZSzLi9bc3IopLh/81efa3fad26d64a726fbdaa06eec09de/10A.png)

__Details of the Program__
 
For Scalien Share, the Workplace Operations team wanted to kick off a series of questions sent to our #social channel on Slack. Some prompts include “What has been your favorite indoor weekend activity?” and “Share a photo of your go-to mug!” 
![Scaliens Share: Favorite TV Show](//images.ctfassets.net/1qgw036lpkfx/2IutT4132A3Aiqfi8OVz0r/4588ed29d74f097e6592328ea46d2da3/10B.png)

__Social Connection__
 
While our design manager [Jeff Hilnbrand](https://www.linkedin.com/in/jhilnbrand/) is taking an intentional break from design work and managing his team, he also has been a facilitator of culture at Scale. We keep our interest group channels engaging with prompts like “What did you cook last week?” in #cooking or “What are you bumping?” in #music. Jeff shares his reflection on social connection through Slack:

Design Manager, Jeff Hilnbrand, shares his experience on Slack prompts.

Jeff Hilnbrand

Jeff Hilnbrand Quote

## Learnings
While transitioning events to a remote setting, we couldn’t always smoothen out every detail before launch of the program. We take some of the areas of improvements as opportunities to learn and improve our future initiatives. 

__Balancing Competition with Participation__ 
Our first wellness challenge emphasized competition over participation. Looking back, in addition to the competitive aspect, we would ideally reward those who were just participating. Focusing on competition only rewards the top individuals. Wellness programs shouldn’t always be about winning or performance. 

__Encouraging grassroots participation__ 
As the people team rallies to foster connection, we could’ve also encouraged employees or given ownership to them to drive some of these initiatives. For example, asking an employee to lead a Scalien share prompt would potentially ensure that these prompts are not always coming from the people team but also inspired cross-functionally from other teams. 

__Finding balance between being structured and unstructured__ 
Lastly, finding the balance between structured connections and organic connections is challenging. For Speed Scaling, we messaged out the questions in timed intervals to keep the group moving to the next question. In retrospect, we realized that some groups want to stay on one question longer than another question. Giving all questions in advance would allow the breakout groups to talk more organically, feeling free to stay on a particular discussion point that they were interested in lingering on. 

Remote events and programs posed new types of challenges, but we’re grateful for Scale giving us the space to experiment ways to keep our employees close during these uncertain times.

Shelter-in-place restrictions posed new challenges that inspired our people team to think of creative ways to keep the team connected.

Fostering connection when everyone is remote

Keeping The Team Close While Distancing

In these unprecedented times, COVID-19 has brought out a renewed and inspiring sense of collaboration in AI and research communities as we work toward solving pressing issues. But the pandemic has also exacerbated some of the difficulties of developing new technologies at scale.
For example, as we shelter in place around the world, the promise of autonomous vehicles (AVs) to improve access to critical goods and services has never felt more <a href="https://mashable.com/article/autonomous-vehicle-perception-coronavirus/" target="_blank">relevant</a>. But as we realize more ways these technologies could improve our lives, the essential data collection and testing that power them have rightly been suspended to ensure the safety of those involved.
That’s why today we’re launching <a href="https://pandaset.org/" target="_blank">PandaSet</a>: a new open-source dataset for training machine learning (ML) models for autonomous driving released in partnership with the LiDAR manufacturer Hesai.
While many AV companies are turning to complementary techniques and simulated data to continue their work, there is often no substitute for high-quality data that captures the complex and often messy reality of driving in the real world.

This image shows the 2D camera view on the left with raw LiDAR view in the top right, LiDAR view with photo colors projected on the point clouds in the middle, and LiDAR view after the point clouds have been semantically segmented. 

LiDAR Comparison

High-quality data is crucial to building safe and effective AV systems. PandaSet is the world’s first publicly available dataset to include both mechanical spinning and forward-facing LiDARs (Hesai’s Pandar64 and PandarGT)—allowing ML teams to take advantage of the latest technologies. It is also the first to be released without any major restrictions on its commercial use.

There are three reasons why we hope AV teams will find PandaSet to be a valuable resource: its content, its quality, and its no-cost commercial license.

### Content

Covering some of the most challenging driving conditions for full level 5 autonomy, PandaSet includes complex urban environments, their dense traffic and pedestrians, steep hills, construction, and a variety of lighting conditions in the day, dusk and evening.

There are more than 48,000 camera images and over 16,000 LiDAR sweeps—more than 100 scenes of 8s each. Capturing sequences in busy urban areas also means there is a high density of useful information, with many more objects in each frame than in other datasets.

### Quality

By combining the strengths of both mechanical spinning and forward-facing LiDARs, PandaSet captures the complex variables of urban driving in rich detail. 

It also includes 28 different annotation classes for each scene as well as 37 semantic segmentation labels for the majority of scenes. With LiDAR data far beyond the capabilities of traditional cuboid labeling, it features Scale’s Point Cloud Segmentation that enables the highest precision and quality annotation of complex objects, such as smoke or rain. 

PandaSet also features Scale’s market-leading [Sensor Fusion](https://scale.com/3d-sensor-fusion) technology, allowing ML teams to blend multiple LiDAR, RADAR and camera inputs into a single point cloud that allows for the semantic segmentation of different objects in LiDAR data. By allowing ML teams to exploit their LiDAR data much more systematically, this makes PandaSet ideal for building highly-performant autonomous systems.

This gif demonstrates the raw LiDAR view, followed by the LiDAR view with photo colors projected on the point clouds, and finally, the scene fully annotated with cuboids. 

LiDAR Flyover

### No-cost commercial license 

Many existing open-source datasets have restrictive licensing terms that allow only research or limited commercial uses. While that can be important for helping ensure data is used appropriately, we wanted to make PandaSet available to the entire community, democratizing access to the latest LiDAR technologies for ML teams around the world at a time when the barriers to data collection are higher.

We all want to accelerate the safe deployment of AVs—and the need for the right data has never been more pressing. By filling the gap for AI and ML developers who might otherwise be unable to build and test new technologies, we hope PandaSet will provide a useful resource for teams building a future for mobility that is safer and more accessible for everyone.

You can find more information about PandaSet [here](https://scale.com/open-datasets/pandaset), with dataset support tools available on [GitHub](https://github.com/scaleapi/pandaset-devkit).


Introducing PandaSet, the world's most advanced LiDAR dataset for commercial use.

Launching the world’s most advanced LiDAR dataset for commercial use.

Introducing: PandaSet. The world's most advanced LiDAR dataset for commercial use

Daniel Lee

<h2 id="introduction">Introduction</h2>
At Scale AI, we label on the order of 10MM annotations per week. To deliver high-quality annotations for this enormous volume of data, we’ve developed a number of techniques including advanced <a href="https://scale.com/blog/point-cloud-segmentation" target="_blank">sensor fusion</a> to provide rich detail about complex environments, <a href="https://scale.com/blog/automatic-segmentation" target="_blank">active tooling</a> to accelerate the labeling process, and <a href="https://scale.com/blog/how-to-label-1m-data-points-week" target="_blank">automated benchmarks</a> to measure and maintain labeler (Tasker) quality. As we work with more customers, more Taskers, and more data, we continue to refine these methods to improve our labeling quality, efficiency, and scalability.
<h2 id="how-we-use-ml">How we use ML</h2>
While this vast quantity of data provides Scale AI with invaluable opportunities to learn and to build upon our annotation processes, it also enables our Machine Learning team to train models that further augment our capabilities. We leverage ML models throughout our annotation pipeline including:
<ol>
<li>Pre-labeling to reduce Taskers’ manual work</li>
</ol>
<figure class="expand-image">
 <img src="//images.ctfassets.net/1qgw036lpkfx/6Ue3M8fPuHkZYtErIIGO54/3f9d1bfa1dd59b376715356ccb349c8e/01.png" />
</figure>
2. Active tooling to maximize Taskers’ speed & efficiency

How We Scale Machine Learning - Video

How Scale AI uses ML to supercharge data annotation pipelines.

Developing scalable ML solutions

How we Scale Machine Learning

Aatish Nayak

Scale AI is pleased to announce the launch of Scale Document - our endpoint for the secure processing of documents.
Advancements in natural language processing (NLP) technologies have spurred initiatives across a wide range of industries to digitize document-based content. Led by the logistics, banking, financial services, and insurance industries, the intelligent document processing (IDP) market is expected to reach $1.1 billion by the end of 2020 according to <a href="https://www.everestgrp.com/2019-06-intelligent-document-processing-market-is-exploding-expected-to-grow-70-80-over-next-2-years-press-release-50323.html" target="_blank">research conducted by the Everest Group</a>.
Scale Document builds on the previously released Scale Text product to better address customers with document processing and understanding use cases with tailored features.
<figure class="expand-image">
 <img src="//images.ctfassets.net/1qgw036lpkfx/4xkmnE7lluFX9ScAfomxQ8/026009856e90c1d74ba613172728fc23/multilineGif.gif" />
 <figcaption>Multi-line transcription: one of many ML-powered tooling features to accelerate human review.</figcaption>
</figure>
Capable of supporting a range of inputs from PDFs, Microsoft Word Documents, JPEGs and more, Scale Document has a few key features:
<ul>
<li>Machine learning to automatically classify and pre-label documents while also accelerating human review with active tooling. In addition to increasing the quality of documents delivered, this also enables the turnaround time for tasks to be as fast as a few hours to meet critical business requirements. (The workflow for how a document is processed is below.)</li>
<li>Anonymization of personally identifiable information (PII) either with name and face blurring or placement of dummy content ensures compliance with data privacy principles.</li>
<li>Flexible and varying levels of workforce security. Customers can select a global or U.S.-based workforce with options to add background checks or annotation in secure facilities depending on requirements. Our global work-from-home workforce also ensures continued delivery of data.</li>
</ul>
<a href="//images.ctfassets.net/1qgw036lpkfx/iT3twJNSTT3AdrYbUlLs3/1e0ee573dfca835b8ddf773cc5010d41/scale_document_flowchart_02.jpg" target="_blank" style="color: inherit;font-weight: normal">
<figure class="expand-image">
 <img src="//images.ctfassets.net/1qgw036lpkfx/iT3twJNSTT3AdrYbUlLs3/1e0ee573dfca835b8ddf773cc5010d41/scale_document_flowchart_02.jpg" />
 <figcaption>The ML + human-in-the-loop workflow that processes documents at high quality and with fast turnaround.</figcaption>
</figure>
</a>
Rigorously tested by a handful of customers during a private beta, Classification, Multilingual Transcription and Data Extraction annotation provide fine-grained understanding of document type, and allows for the ability to extract and understand document content and generate responses based off of document content, such as approving invoices or flagging applications for human review.
Take a look at the <a href="https://scale.com/document" target="_blank">Scale Document product page</a> or <a href="https://scale.com/sales" target="_blank">contact us</a> to talk about your business needs.

Scale Document is the secure endpoint for document processing.

Introducing Scale AI's latest product: Scale Document

Now Available: Scale Document

<h2 id="notice-any-changes-to-our-website%3F">Notice any changes to our website?</h2>
The Scale AI team is excited to introduce a new look and new structure for our product line-up. Scale AI was founded almost four years ago because we believed that high quality training data is the key bottleneck to the development of AI applications. Thanks to feedback from incredible customers, we have significantly expanded our data annotation products over the years to support an increasing range of data inputs and annotation types for computer vision and natural language (NLP) applications.
As we expanded our data annotation capabilities, it became clear that we needed a better way to group and name our products to reflect both the current state and future vision of our product offering.
<h2 id="what-exactly-are-the-changes%3F">What exactly are the changes?</h2>
Sensor Fusion Cuboids and Sensor Fusion Segmentation is now one product called <a href="https://scale.com/3d-sensor-fusion" target="_blank">Scale 3D Sensor Fusion</a>.
<figure class="expand-image">
 <img src="//images.ctfassets.net/1qgw036lpkfx/2houGzSkeiXHADZcVUDpnP/5ae8d3d02a5070c4e11e06f9a86c758b/01__1_.jpg" />
</figure>
3D Cuboids, Bounding Boxes, Image Categorization, Lines &amp; Splines, Polygons and Semantic Segmentation are also now one product called <a href="https://scale.com/image" target="_blank">Scale Image</a> with an addition of Point Annotation.
<figure class="expand-image">
 <img src="//images.ctfassets.net/1qgw036lpkfx/2j1HUM81z13OIbbRCay4aL/befc6c0d35ff7df66838857a7b6b939d/scale_image.jpg" />
</figure>
Video Annotation is now <a href="https://scale.com/video" target="_blank">Scale Video</a>, and has been expanded to include classification, cuboid and polygon annotation in addition to the existing bounding box annotation.
<figure class="expand-image">
 <img src="//images.ctfassets.net/1qgw036lpkfx/2z9kBh22olI5FoHSCODEWr/515387382df67ded8bc00662d547ff11/03__1_.jpg" />
</figure>
Our Natural Language products have been updated to <a href="https://scale.com/text" target="_blank">Scale Text</a> with Scale Audio currently in private beta.
<figure class="expand-image">
 <img src="//images.ctfassets.net/1qgw036lpkfx/5CuqCheg6WwhnbcUM5bRWz/b56672b4c1eaa65977caede62e09f49a/04__1_.jpg" />
</figure>
<figure class="expand-image">
 <img src="//images.ctfassets.net/1qgw036lpkfx/1RqGrBMgSLmIaG6lIZmfhu/6598f6c94f14303594c6ccbde3a2c780/05.jpg" />
</figure>
You may also wonder, “how does this product restructure affect me?” If you are one of our amazing customers, from a practical standpoint, impact to you is minimal. While the product names are different, the products themselves are still the same, and we will continue to deliver data at the quality you expect. If you’re learning about Scale AI for the first time, we hope this rebrand allows you to better understand our full product suite and its capabilities.
<h2 id="what%E2%80%99s-next%3F">What’s Next?</h2>
The Scale AI team remains committed to delivering products customers love and rely on. We are grateful for all of our incredible and innovative customers for their support. In the coming weeks and months, we’ll be announcing and rolling out new products and additional annotation types to existing products.
In the meantime, if you have any questions about todays’s announcement, reach out to us at <a href="mailto:support@scale.com" target="_blank">support@scale.com</a>. Otherwise, we invite you to explore our updated product line up.
<figure class="expand-image">
 <img src="//images.ctfassets.net/1qgw036lpkfx/4WVGAESruBGqfVBitOG0FX/3be113efdad428155b3cce7b86e373f4/06.jpg" />
</figure>

Explore the new product line-up: Scale 3D Sensor Fusion, Scale Image, Scale Video and Scale Text

Introducing a new look and new structure for our core product line-up

Introducing a New Look for Scale AI Products

Andrew Liu

Machine learning is a field of study with tremendous strides being made through active academic research. The machine learning team at Scale AI regularly hosts reading groups to discuss papers they find interesting to give us context in a rapidly evolving, research-heavy field. In this blog post, we go over the papers the team has read throughout the quarter and provide insights on how a paper influenced our own work here at Scale AI when relevant.
<h3 id="the-lottery-ticket-hypothesis%3A-finding-sparse%2C-trainable-neural-networks"><a href="https://arxiv.org/abs/1803.03635" target="_blank">The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks</a></h3>
Authors: Jonathan Frankle, Michael Carbin 
Presenter: <a href="https://www.linkedin.com/in/ppyht2/" target="_blank">Hongbo Tian</a>
Pruned neural network architectures created by removing unnecessary weights from trained dense neural networks can attain various benefits without sacrificing accuracy. For example, smaller architectures have lower storage requirements, and certain hardware architectures can take advantage of sparsity, leading to massive computational performance gains at inference time. However, training the pruned network from scratch (as opposed to training the full network and pruning) has historically achieved poor accuracy.
This paper presents the hypothesis that dense neural networks will contain some subnetwork that can reach the same performance of the original network when trained in isolation, and the authors cover some techniques to uncover the “winning subnetwork.” Larger network architectures have typically been considered superior, but this paper demonstrates that such a belief is not necessarily true. The lottery ticket conjecture matters when developing machine learning models here at Scale AI, since we know that we can build networks that tackle ambitious challenges without being blocked on unreasonable monetary or compute requirements.
<h3 id="mastering-atari%2C-go%2C-chess-and-shogi-by-planning-with-a-learned-model"><a href="https://arxiv.org/abs/1911.08265" target="_blank">Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model</a></h3>
Authors: Julian Schrittwieser, Ioannis Antonoglou, Thomas Hubert, Karen Simonyan, Laurent Sifre, Simon Schmitt, Arthur Guez, Edward Lockhart, Demis Hassabis, Thore Graepel, Timothy Lillicrap, David Silver 
Presenter: <a href="https://www.linkedin.com/in/ppyht2/" target="_blank">Hongbo Tian</a>
<figure class="expand-image">
 <img src="//images.ctfassets.net/1qgw036lpkfx/7tRJ5CkMZZu62uSNJnkmAF/db622190e14f2b667705f186a6a209cf/scale_blog_ml_papers_cover_01.jpg" />
 <figcaption>MuZero’s performance, compared to the state of the art. After sufficient training, MuZero’s Elo (denoted by the blue lines) surpasses that of AlphaZero for boardgames and R2D2 in Atari.</figcaption>
</figure>
Games like chess and Go tend to have deterministic and well-defined rules. Such games aren’t representative of real world problems because most realistic problems occur in an environment with complex or unknown rules. This paper presents MuZero, a successor to AlphaZero that matches AlphaZero’s performance in Go but can learn to play Atari games, all without any prior knowledge to the game rules. MuZero builds off of AlphaZero’s search algorithms, but it also uses a learned environment model when training to capture aspects of the future that are relevant for planning.
Although MuZero don’t know the game rule or environmental dynamics of the challenges it faces, it matches AlphaZero in logically complex games and beats the state-of-the-art model-free reinforcement learning algorithms in visually complex games (e.g. Atari games). Such a feat makes MuZero a pioneer in developing AI to solve realistic problems where there’s no perfect simulator and demonstrates the feasibility of applying learning methods to more ambiguous problems.
<h3 id="fixmatch%3A-simplifying-semi-supervised-learning-with-consistency-and-confidence"><a href="https://arxiv.org/abs/2001.07685" target="_blank">FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence</a></h3>
Authors: Kihyuk Sohn, David Berthelot, Chun-Liang Li, Zizhao Zhang, Nicholas Carlini, Ekin D. Cubuk, Alex Kurakin, Han Zhang, Colin Raffel 
Presenter: <a href="https://www.linkedin.com/in/felix-lau-0386641a/" target="_blank">Felix Lau</a>
<figure class="expand-image">
 <img src="//images.ctfassets.net/1qgw036lpkfx/6Bsf1WLDq5PaSaBH0axOxd/b1d43cc489d75ad80f0be1fb207bbd46/scale_blog_ml_papers_cover_02.jpg" />
 <figcaption>The general architecture of FixMatch. The model is trained to make predictions on a strongly-augmented version of an unlabeled image, based off the prediction results of a weakly-augmented version of the same image.</figcaption>
</figure>
The authors combine two common semi-supervised learning methods, consistency regularization and pseudo-labeling, to create a new algorithm that achieves state-of-the-art performance on standard, semi-supervised learning benchmarks. For unlabeled data, the algorithm applies both strong augmentation and weak augmentation. If the confidence for unlabeled data is higher than a certain threshold, the algorithms apply a cross-entropy loss between the prediction from weakly and strongly augmented images.
<h3 id="reformer%3A-the-efficient-transformer"><a href="https://arxiv.org/abs/2001.04451" target="_blank">Reformer: The Efficient Transformer</a></h3>
Authors: Nikita Kitaev, Łukasz Kaiser, Anselm Levskaya 
Presenter: <a href="https://www.linkedin.com/in/ppyht2/" target="_blank">Hongbo Tian</a>
<figure class="expand-image">
 <img src="//images.ctfassets.net/1qgw036lpkfx/6lnMtUYpSWo3TKYEccjyMj/477f292cd879424b71dfb59d534a79bb/scale_blog_ml_papers_cover_03_B.jpg" />
 <figcaption>A visual depiction of attention with locality sensitive hashing. By hashing and then bucketing keys, attention doesn’t need to handle all pair-wise keys.</figcaption>
</figure>
Training good transformers can be prohibitively expensive: the largest version of GPT-2 contains 1.5 billion parameters, while NVIDIA’s Megatron-LM has up to 8.3 billion parameters. The authors of Reformer present a model that performs on par with preceding transformers while being more memory-efficient and faster to train. Due to the decreased resource requirements, Reformer makes Transformers more widely available to use in practice while also being able to handle longer sequences in general.
Rather than using the standard dot-product attention, Reformer uses a less expressive attention mechanism combined with locality sensitive hashing, which reduces the complexity of attention. The classical self-attention mechanism attends to all pair-wise elements in the input sequence, giving it quadratic algorithmic complexity. By reducing the expressiveness of self-attention, Reformer can leverage locality-sensitive hashing via bucketing similar elements together, thus simplifying self-attention to linearithmic complexity. This mechanism ultimately enables Reformer to process the entirety of Crime and Punishment in a single training example.
<h3 id="the-case-for-learned-index-structures"><a href="https://arxiv.org/abs/1712.01208" target="_blank">The Case for Learned Index Structures</a></h3>
Authors: Tim Kraska, Alex Beutel, Ed H. Chi, Jeffrey Dean, Neoklis Polyzotis 
Presenter: <a href="https://www.linkedin.com/in/ppyht2/" target="_blank">Hongbo Tian</a>
<figure class="expand-image">
 <img src="//images.ctfassets.net/1qgw036lpkfx/1eRxG0ClufEVFJH47Ios8T/4846e6bec241f70b29dbcb1b23585a23/scale_blog_ml_papers_cover_04.jpg" />
 <figcaption>A comparison of a traditional and learned hash map. Learned data structures replace the hashing function with a model that achieves better performance.</figcaption>
</figure>
Many classic data structures (such as B-Trees, Hash maps, and Bloom filters) rely on some index structure to map a key to a position of a record. However, index structures can be generalized as a model that takes in a key and infers information about a data record. The authors explore replacing traditional index structures with deep-learning models, and their results demonstrate that machine learning can greatly increase speed of classic data structures that have already been optimized over decades. This paper is novel in the sense that it exposes opportunities for large performance improvements in data structures that have already been optimized for decades. It also serves as a reminder to our team at Scale AI, that excellence can exceed that of our initial expectations and that machine learning can have very non-obvious use cases.
<h3 id="pointrend%3A-image-segmentation-as-rendering"><a href="https://arxiv.org/abs/1912.08193" target="_blank">PointRend: Image Segmentation as Rendering</a></h3>
Authors: Alexander Kirillov, Yuxin Wu, Kaiming He, Ross Girshick 
Presenter: <a href="https://www.linkedin.com/in/felix-lau-0386641a/" target="_blank">Felix Lau</a>
<figure class="expand-image">
 <img src="//images.ctfassets.net/1qgw036lpkfx/64iMDLEBEkU50sNifZonox/b754aa427c639f3921af92c4416a5ce7/image__18_.png" />
 <figcaption>Inference results on a scene with and without PointRend. PointRend is able to handle the edges, such that it properly segments the side mirrors of a car.</figcaption>
</figure>
Image segmentation can suffer in accuracy around the borders of various classes and difficult regions. As a result, the authors of this paper create a neural network module call PointRend that adaptively runs point-based segmentation in higher resolutions at selected locations, leading to crisp boundaries compared to previous segmentation techniques. PointRend can be applied on top of existing instance and semantic segmentation tasks. At Scale AI, we’ve found that getting the edges between classes right is the most critical when it comes to producing high quality labels for 2D segmentation.
<h3 id="rangenet%2B%2B%3A-fast-and-accurate-lidar-semantic-segmentation"><a href="http://www.ipb.uni-bonn.de/wp-content/papercite-data/pdf/milioto2019iros.pdf" target="_blank">RangeNet++: Fast and Accurate LiDAR Semantic Segmentation</a></h3>
Authors: Andres Milioto, Ignacio Vizzo, Jens Behley, Cyrill Stachniss 
Presenter: <a href="https://www.linkedin.com/in/danielhavir/" target="_blank">Daniel Havir</a>
<figure class="expand-image">
 <img src="//images.ctfassets.net/1qgw036lpkfx/1AbcHKTLDFXylf3LUU3EiX/7241304eda6b5e8b9aa710c0b17a4d14/scale_blog_ml_papers_cover_05.jpg" />
 <figcaption>Diagram of the various states of data representation during LiDAR segmentation. Each letter corresponds to a step in the approach that RangeNet++ takes.</figcaption>
</figure>
High-level semantic perception tasks in autonomous vehicles are primarily solved through high resolution cameras, even though there are a wider variety of sensor modalities available for autonomous vehicles. At Scale AI, we are familiar with the problems defined in this paper. The paper details an approach to improving LiDAR-only semantic segmentation. The authors use a 4-step approach:
<ol>
<li>Reduce the LiDAR point cloud to two-dimensional range images that encode the Euclidean distance along with the original coordinates and LiDAR remission.</li>
<li>Run the range image representation through a 2D semantic segmentation model.</li>
<li>Project the model output back to the original point cloud format.</li>
<li>Post-process the output using a novel KNN approach to account for occlusion errors introduced with the 2D projection.</li>
</ol>
<h3 id="fast-differentiable-sorting-and-ranking"><a href="https://arxiv.org/abs/2002.08871" target="_blank">Fast Differentiable Sorting and Ranking</a></h3>
Authors: Mathieu Blondel, Olivier Teboul, Quentin Berthet, Josip Djolonga 
Presenter: <a href="https://www.linkedin.com/in/chiao-lun-cheng-87a4a34/" target="_blank">Chiao-Lun Cheng</a>
<figure class="expand-image">
 <img src="//images.ctfassets.net/1qgw036lpkfx/2ImbW0lRPiVSsieOkY0t5k/1c58b880baf4bceb2531b9fe9659c927/scale_blog_ml_papers_cover_06.jpg" />
 <figcaption>Illustration of the soft sorting and ranking operators.</figcaption>
</figure>
Sorting and ranking are basic and common operations in computer science and machine learning. However, due to their natural properties, they are not differentiable. The authors propose the first differentiable sorting and ranking operations, while maintaining their time and space complexity. To reduce sorting and ranking to a differentiable form, the authors construct the operators as projections onto a permutohedron, take the convex hull of the ordering permutations, and convert the results to an isotonic optimization problem.
If you’re interested in joining our growing engineering team, take a look at our <a href="https://scale.com/careers" target="_blank">careers page</a> for open positions.

The machine learning research papers the Scale AI team read and discussed in Q1 2020.

Research Papers the Scale AI Team has Read in Q1 2020

Scale AI Machine Learning Digest - Q1 2020

As the COVID-19 pandemic continues to spread, it affects all of our communities in different ways. The health and safety of Scale AI’s employees, as well as our global community of labelers, is a critical matter and a responsibility we do not take lightly. That’s why we’ve taken action over the past several weeks to protect both our full-time and contract employees, and ensure we can continue to offer the high-quality services our customers expect.
We also want Scale AI to make a contribution to combating this disease if we can. Research to track and contain the virus is in its early days, but we know that better data could make a difference. To help these efforts, Scale AI will provide free data annotation services for any researcher working on diagnosing, containing, tracking or finding a vaccine or long-term cure for the virus. If you’re working on a project, you can reach out to us <a href="https://events.scale.com/accelerating-covid-19-research" target="_blank">here</a>.
<h2 id="scale-ai%E2%80%99s-employees%2C-labeling-community-and-customers">Scale AI’s Employees, Labeling Community and Customers</h2>
The safety of our team, both full-time and contractors, is a top priority. As a technology company, we are fortunate to be able to leverage existing communications tools to maintain collaboration. Scale AI’s ML-enabled platform also gives a global, work-from-home workforce the ability to deliver data annotation at the level of quality and volume our customers expect. Anyone working with a basic laptop and internet connection can utilize our platform. We are proud to provide a safe option for workers to continue to earn wages to support themselves and their families throughout this period of uncertainty.
We have also reached out to our customers to let them know how we’re working to deliver products and services. We do not expect any disruption to our work at the moment, however, because the situation remains fluid, we will continue to closely monitor our operations. In the event of a disruption, we will make sure to coordinate with your teams to minimize any impact.
<h2 id="accelerating-covid-19-research">Accelerating COVID-19 Research</h2>
Adjusting to the reality of a new disease has been difficult and unsettling, but it has been heartening to see a range of companies and organizations come together to help combat this pandemic. We want to make our own contribution to these efforts where we can.
Given the scale of the pandemic, researchers should have every tool at their disposal as they try to track and counter this virus. Researchers have <a href="https://deepmind.com/research/open-source/computational-predictions-of-protein-structures-associated-with-COVID-19" target="_blank">already shown</a> how new machine learning techniques can help shed new light on this virus. But as with all new diseases, this work is much harder when there is <a href="https://www.wsj.com/articles/coronavirus-reveals-limits-of-ai-health-tools-11582981201" target="_blank">so little existing data to go on</a>.
In those situations, the role of well-annotated data to train models or diagnostic tools is even more critical.
Our mission at Scale AI is to accelerate the development of AI applications. To help in those efforts, Scale AI will provide free data annotation services for any researcher working on diagnosing, containing, tracking or finding a vaccine or long-term cure for the virus. Advancing and accelerating health research and science is critical to reducing the spread of viral infections. If you have data labeling needs related to this research, please let us know <a href="https://events.scale.com/accelerating-covid-19-research" target="_blank">here</a>.

Scale AI will offer free labeling to researchers working to contain the spread of the virus.

Scale AI is committed to containing the spread and being a resource to researchers working on COVID-19.

Supporting researchers and our community during COVID-19

Self-driving cars promise to improve mobility, build a more efficient transportation system and even save lives.
But while the industry has been making great progress, to date the field has struggled to build cars and artificial intelligence models that can handle all weather conditions.
Snow is hard to drive in—as many drivers are well aware. But wintry conditions are especially hard for self-driving cars because of the way snow affects the critical hardware and AI algorithms that power them.
Take a car trained to drive on the wide, sunny boulevards of Arizona or California and place it on a snowy street like this, and things will go wrong:
<figure class="expand-image">
 <img src="//images.ctfassets.net/1qgw036lpkfx/1vIW2cAbNfsASiyzMRG5lX/15ecbabdef4f718cfd44835e9451dec6/Screen_Shot_2020-01-10_at_1.23.17_PM.png" />
</figure>
Snow covering the road markings would make it hard for the car to stay in its lane.
Accumulated snow significantly reduces the color contrast, making it harder for the car’s system to recognize objects, like trees or parked cars covered in snow—or even pedestrians.
Most problematically, snowfall makes it hard for a vehicle’s cameras and LiDAR sensors to see, which significantly reduces safety:
<figure class="expand-image">
 <img src="//images.ctfassets.net/1qgw036lpkfx/7HU58f8zXZyY3nbB8JXBfy/56aef1a88523676ef6cacaf184749677/scale-waterloo-comparison.png" />
</figure>
A skilled human driver can handle the same road in all weathers—but today’s AV models can’t generalise their experience in the same way. To do so, they need much more data.
Until now, there has not yet been an open dataset of annotated LiDAR data and images of driving in snowy weather that researchers can use to develop more robust systems.
<h2 id="cadc%3A-a-new-dataset-for-wintry-conditions">CADC: A new dataset for wintry conditions</h2>
In partnership with the University of Waterloo and the University of Toronto, we’ve been applying our technology to this challenge.
Today, we’re open sourcing the world’s first AV dataset in wintry conditions—the Canadian Adverse Driving Conditions dataset (CADC, or cad-see.)
Drawn from 20 km of driving in harsh, snowy conditions in southwestern Ontario over the past two winters, CADC contains 7,000 frames of LiDAR and camera data spread over 70 driving sequences—featuring a wide range of perception challenges for AV models that can handle wintry conditions. Scale provided the annotations that makes the data usable by researchers.
Researchers at the Universities of Waterloo and Toronto have also released a detailed academic paper outlining the technical features of the data collected and object detection labels provided, which you can view on arXiv <a href="https://arxiv.org/abs/2001.10117" target="_blank">here</a>.
Ultimately, this is about safety. Without robust data, self-driving models can’t learn how to handle snowy conditions—preventing us from being able to make AV systems a reality for the many millions of people who live in places that receive wintry weather.
As Professor Krzyzstof Czarnecki at the University of Waterloo says, “We want to engage the research community to generate new ideas and enable innovation. This is how you can solve really hard problems, the problems that are just too big for anyone to solve on their own.”
We hope this dataset will make it easier for researchers to achieve the next-generation breakthroughs in AV systems that will take us one step closer to self-driving vehicles that work safely in a wide range of different environments.
As Professor Steven Waslander at the University of Toronto says, “We’re hoping that both industry and academia go nuts with it. We want the world to be working on driving everywhere.
“Bad weather is a condition that is going to happen.”
You can find the dataset on the <a href="http://cadcd.uwaterloo.ca/" target="_blank">Waterloo website</a>, with dataset support tools available on <a href="https://github.com/mpitropov/cadcd_devkit" target="_blank">GitHub</a>.

Open source autonomous vehicle dataset for wintry conditions.

Overcoming a critical challenge to the widespread use of autonomous vehicles

Open Sourcing the World's First AV Dataset for Wintry Environments

Akshat Bubna

OpenAI recently published <a href="https://openai.com/blog/fine-tuning-gpt-2/" target="_blank">a paper on fine-tuning GPT-2</a>, where they used Scale AI to collect the preferences of human labelers to improve their language models. Although we were already labeling millions of text and computer vision tasks per day at the time, the unique latency requirements and subjective nature of OpenAI’s tasks posed a new challenge for us. In particular: how do you scalably maintain the quality of labels, without having labelers check each other’s work? Today we’re sharing a deep dive into our approach to the problem, the automatic benchmark mining system we built to solve it, and the things we learned along the way. In doing so, we hope to illustrate some of the many challenges that make scalable data labeling such an interesting area of work.

## The Problem

First, let’s take a look at one of these tasks. Given a snippet of text and some possible continuations for it, the researchers at OpenAI wanted humans to pick the continuation that best fit a given prompt, such as “which continuation has the most positive sentiment” or “which continuation is the most descriptive”. 

How to Label 1M Data Image 01

But instead of labeling this data in a big batch like we usually do (*offline*), OpenAI wanted to put their model in a closed loop with our labelers, so the data was labeled *online*: the model would generate some text samples, have the humans rate it using our API, train based on human preferences, and then repeat this process over a few days.

How to Label 1M Data Image 02

How to Label 1M Data Image 03

How do you scalably maintain the quality of labels, without having labelers check each other’s work? Take a deep dive into how we solved this problem while working with OpenAI on fine tuning their GPT-2 model.

Resources

The Learning Curve: Fireside Chat with Emily Denton

The Learning Curve

The Learning Curve: Applying Game Rating Systems to Data Labeling

Get Labeled Data Today