I am interested in enhancing human performance and longevity using technology.
At Insane AI, we are using motion tracking using computer vision to build the future of fitness, allowing your body to be the game controller in an immersive fitness game.
My research has revolved around computer vision, biomedical sensor systems, and pedestrian understanding for autonomous driving.
Work
Insane AI
Freeing humanity from the tyranny of Zoom fitness classes.
Pedestrian understanding for fully autonomous driving
Detection, finer understanding, and path prediction of pedestrians and cyclists on the road is a non-negotatiable pre-requisite for allowing algorithms to control vehicles in unconstrained scenarios. I worked on the perception problem and was responsible for end-to-end delivery of the project milestones.
Medd
Precise, effective medical treatment can be delivered once a timely and accurate diagnosis is in place. I co-founded Medd in college to simplify the process of getting lab tests at home, with quick and insightful digitized reports. In June 2016, Medd was acquired by 1mg Technologies.
Research
VRU Pose-SSD: Multiperson Pose Estimation For Automated Driving
We present a fast and accurate method for joint detection and pose estimation of pedestrians and cyclists, for autonomous driving usecases.
Learning 3D Human Pose from Structure and Motion
3D human pose estimation from a single image is a challenging problem, especially for in-the-wild settings, due to the difficulty of annotating depth. We propose two anatomically inspired loss functions, and use them with a weakly-supervised learning framework, to jointly learn from large-scale in-the-wild 2D and indoor/synthetic 3D data.