2025
SAM 3D
Evaluating a new state-of-the-art model for human 3D pose estimation for lifting form analysis.
Perfect Form: To the Cloud
With a functioning local prototype of my powerlifting form analysis application, I was ready to migrate the application to the cloud. Could I transform my demo to a real hosted application leveraging cloud GPUs with a UI capable of displaying all of the extracted metrics?
2023
Perfect Form: Computer Vision for Powerlifting
As an avid powerlifter looking for a problem to learn machine model deployments, I gravitated towards designing and implementing a microservice architecture that would leverage the latest computer vision models to analyze powerlifting from through inferred biomechanical metrics. How far could I push this idea to build a useful lifting tool for my training?
2020
Lost Cities Part 2: Learning from Deep Q-Networks
Now that we developed a deep Q-network that achieved a mean score of 45 points in Lost Cities, what can we learn from it? We probe this network to understand the high-level strategy choices it makes, and we quantify the distributions of important metrics governing its strategy. To better evaluate its performance, we want to play against the network ourselves. Can we train an object detection architecture so that we can play against the deep Q-network using a game with physical cards?
