Papers
9
Total Citations
172
H-Index
6
About
Chahat Deep Singh is a pioneering roboticist whose research lies at the intersection of minimal perception, active vision, and tactile sensing for resource-constrained autonomous systems. His work fundamentally challenges the assumption that robots require heavy, power-intensive sensors to achieve intelligent behavior. Singh’s most influential contribution, "GapFlyt" (90 citations), introduced an active vision-based, structure-less gap detection method for quadrotors, demonstrating that aerial robots can navigate complex environments using minimalist perceptual strategies rather than building full 3D maps. He further advanced this philosophy through "Ajna," a framework for generalized deep uncertainty estimation on parsimonious robots, and "PRGFlow," which developed SWAP-aware optical flow for aerial navigation. In tactile robotics, his "AcTExplore" (13 citations) pioneered active tactile exploration for unknown object understanding, while "NudgeSeg" showed how repeated physical interaction enables zero-shot object segmentation. Singh’s recent work on microsaccade-inspired event cameras (2024) bridges biological vision and neuromorphic sensing for high-dynamic robotics. His research consistently addresses the critical challenge of enabling autonomy under extreme Size, Weight, Area, and Power (SWAP) constraints, making him a leading voice in the future of deployable, efficient robotic systems for real-world applications.
Research Focus
Key Achievements
Top Papers
- 1
- 2Microsaccade-inspired event camera for robotics25 citations · 2024
- 3
- 4AcTExplore: Active Tactile Exploration on Unknown Objects13 citations · 2024
- 5
- 6NudgeSeg: Zero-Shot Object Segmentation by Repeated Physical Interaction7 citations · 2021
- 7PRGFlow: Benchmarking SWAP-Aware Unified Deep Visual Inertial Odometry6 citations · 2020
- 8WorldGen: A Large Scale Generative Simulator5 citations · 2023
- 9Minimal perception: enabling autonomy in resource-constrained robots2 citations · 2024