Makar Tapaswi
Papers
1
Total Citations
6
H-Index
1
About
Makar Tapaswi is a leading researcher in computer vision and robotics, with a focus on learning manipulation skills from visual demonstrations. His work bridges the gap between video understanding and physical interaction, enabling robots to acquire object manipulation abilities by watching single video examples. In his notable 2022 paper, "Learning Object Manipulation Skills from Video via Approximate Differentiable Physics," Tapaswi introduced an optimization framework that reconstructs coarse, temporally evolving 3D scenes from demonstration videos, allowing robots to mimic complex actions without extensive training data. This approach, while still early in its citation impact (6 citations), represents a significant step toward zero-shot robotic learning. Tapaswi's broader contributions include advancing differentiable physics engines for vision-based control, making him a key figure in the emerging field of video-to-robot transfer learning. His work is particularly influential for researchers exploring how to reduce the data and engineering burden in robotic skill acquisition, with potential applications in manufacturing, healthcare, and domestic assistance.
Research Focus
Key Achievements
Top Papers
- 1