Solving communications puzzles in a high functioning environment to visualize complicated advancements in artificial intelligence and engineering research.
Egocentric 4D perception
Egocentric 4D Perception (EGO4D) is a massive-scale, egocentric dataset and benchmark suite collected across 74 worldwide locations and 9 countries, with over 3,025 hours of daily-life activity video. My role was to produce video and graphic deliverables to introduce the open source tool. Read the blog at Meta Ai.
Inside the labs, 2020
ReSkin: Tactile sensing
ReSkin is a new open source touch-sensing “skin” created by Meta AI researchers, in collaboration with Carnegie Mellon University, that can help researchers advance their AI’s tactile-sensing skills quickly and at scale. In this blog post for Meta Ai I developed all visual assets in collaboration with researchers to introduce the new technology through video. The challenge was to sift through existing assets from researchers to make a seamless informational experience to communicate the technology behind the product that accurately reflected what they had been working on. Read more on Meta Ai.
Storyboard
Meta Ai Few-Shot Learner
AI needs to learn what to look for and it typically takes several months to collect and label thousands, if not millions, of examples necessary to train each individual system to spot a new type of content. Meta has created a new Ai system Few-Shot Learner (FSL) that can adapt to take action on new or evolving types of harmful content within weeks instead of months. In this blog I was the designer for visual assets using the brand’s guidelines to communicate the efficiency of the new method. Read more on Meta AI | Read more on FB news
How machine learning powers facebook’s News Feed ranking algorithm
An under-the-hood look into how machine learning (ML) decides what goes into people’s News Feeds through the Facebook app. In this project I worked with researchers and engineers to develop infographics while using the AI blog brand to match and the accompany video in explaining how the data is parsed through a closer lens for Engineering @Meta. Read the blog post.
In the video itself I was a new designer at the time who assisted in the hand off of animation files, and collated the style with the Engineering brand for collateral deliverables.
Expire Span blog
In the accompanying blog post, I gave form to abstract ideas in artificial intelligence to give an in depth look into how Expire Span gave neural networks the ability to forget data at scale similarly to how human memory works, using Meta’s Ai brand guides with my own touch.
Probablistic Flakiness: How do you test your tests?
In this introduction to a new tool by researchers, I collated large amounts of data that the program generated to create informational graphics and animations suitable for industry professional to understand how it works. The challenge here was to balance how the data would appear to tech industry journalists and professionals while maintaining the integrity of the data. Read more on Engineering at Meta.