Viraj Mehta
Papers
4
Total Citations
242
H-Index
4
About
Viraj Mehta’s research sits at the intersection of robotic manipulation, computer vision, and deep learning, with a central focus on enabling robots to reason about tasks rather than simply grasping objects. His most influential work, “Learning Task-Oriented Grasping for Tool Manipulation from Simulated Self-Supervision,” has accumulated over 185 citations and addresses a critical gap in robotics: while traditional grasping prioritizes stability, Mehta’s approach teaches robots to grasp tools based on the desired functional outcome—such as hammering or cutting—by leveraging simulated self-supervision. This paradigm shift from task-agnostic to task-oriented grasping has significant implications for autonomous systems performing complex, real-world tasks. In addition, Mehta contributed to 3D shape reconstruction with “DeformNet,” a free-form deformation network that efficiently reconstructs 3D shapes from a single image, offering computational advantages over voxel- or point-cloud-based methods. His work demonstrates a consistent commitment to bridging perception and action, making him a notable figure in the robotics and AI community.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4