cv

BS

University of Missouri

Biomedical Engineering

MS,PhD

Northwestern University, Shirley Ryan AbilityLab

Biomedical Engineering

Graduate Research Fellowship

National Science Foundation

T32 Training Grant

National Institutes of Health

Differentiable Biomechanics for Markerless Motion Capture in Upper Limb Stroke Rehabilitation: A Comparison with Optical Motion Capture

arXiv

Validates our computer vision approach to gold standard biomechanics in the upper limb.

Biomechanical Arm and Hand Tracking with Multiview Markerless Motion Capture

2024 10th IEEE RAS/EMBS International Conference for Biomedical Robotics and Biomechatronics (BioRob)

Applies earlier joint optimization of biomechanical models to the ARMS Lab's leading hand model.

Fusing Uncalibrated IMUs and Handheld Smartphone Video to Reconstruct Knee Kinematics

2024 10th IEEE RAS/EMBS International Conference for Biomedical Robotics and Biomechatronics (BioRob)

Presents a deep learning approach to fit a biomechanical model to video and IMU data collected during walking.

Hyperpolarized 129Xe MRI, 99mTc scintigraphy, and SPECT in lung ventilation imaging: a quantitative comparison

Academic Radiology

Compares a recently FDA approved method for lung imaging to the clinical standard in a large patient cohort.

Optimizing Trajectories and Inverse Kinematics for Biomechanical Analysis of Markerless Motion Capture Data

2023 International Conference on Rehabilitation Robotics (ICORR)

Develops a novel approach to jointly optimize body scale and joint paramaters for markerless motion capture.

Self-Supervised Learning of Gait-Based Biomarkers

Predictive Intelligence in Medicine

Creates a transformer that given timeseries of gait data learns a self-supervised embedding space useful for downstream tasks.

Enhanced selectivity of transcutaneous spinal cord stimulation by multielectrode configuration

Journal of Neural Engineering

Explores electrode placement and configuration to improve the selectivity of transcutaneous spinal cord stimulation.

Markerless Motion Capture and Biomechanical Analysis Pipeline

arXiv

Hardware and software pipeline for markerless motion capture and biomechanical analysis.

Deep reinforcement learning for global trajectory tracking

Incorperating phyiscal constrains such as forces and torques enable us to create physically grounded reinforcement learning policys to more accurately model movement in global space and measure joint torques.

• reinforcement learning
• computer vision
• biomechanics

Single camera human pose estimation

Accurately recoving human pose from a single camera could democratize access to biomechanical analysis and enable big data studies of human movement. My algorithms incorperate physical constraints to improve accuracy and interpretability.

• Computer Vision
• Deep Learning
• Biomechanics

Video-IMU fusion for human pose estimation

While video is a powerful and accessible sensor, it will always be limitied by occlusions. This work combines video and IMU data to improve accuracy and robustness of human pose estimation.

• Computer Vision
• Deep Learning
• Sensor Fusion

Speaker

National Biomechanics Day

Developed interactive sessions for high school students to learn about biomechanics, prosthetics and computer vision.

• Awarded 'Climate Hero' award by Greenpeace for my efforts organizing the march.
• Men of the year 2014 by Time magazine