B. Ravichandran
Papers
2
Total Citations
12
H-Index
2
About
B. Ravichandran’s research lies at the intersection of computer vision, biomechanics, and deep learning, with a core focus on inferring human dynamics from visual data. His major contribution is pioneering end-to-end deep learning architectures—PressNet and PressNet-Simple—that directly regress 2D foot pressure heatmaps and center of pressure from video frames of human motion. This work bridges the gap between kinematic pose estimation and dynamic force analysis, enabling non-invasive, video-based stability assessment. His most cited paper (2020, 7 citations) validates these models for estimating base of support, while his foundational 2018 work (5 citations) established the framework for learning dynamics from kinematics. Though early in his career, Ravichandran’s approach has significant implications for kinesiology, rehabilitation, and robotics, where understanding postural control and gait is critical. By replacing costly force plates with camera-based inference, his work opens new avenues for accessible biomechanical analysis. His research is particularly notable for tackling the challenging problem of predicting physical forces from purely visual input—a step toward truly understanding human movement from video alone.
Research Focus
Key Achievements
Top Papers
- 1
- 2