Sudharshan Suresh
Papers
5
Total Citations
152
H-Index
4
About
Sudharshan Suresh is a robotics researcher specializing in tactile perception, multimodal sensing, and robot manipulation, with a particular focus on enabling machines to interact with and understand objects in unstructured environments. His work bridges the gap between touch and vision, developing systems that allow robots to build rich spatial awareness during physical interaction. Suresh's most influential contribution, "NeuralFeels with Neural Fields" (2024, 66 citations), leverages neural field representations to fuse visuotactile signals for real-time object pose and shape estimation during in-hand manipulation — a critical step toward human-level robotic dexterity. His complementary work on "ShapeMap 3-D" (2022, 50 citations) introduced an efficient pipeline for 3D shape reconstruction by combining dense tactile sensing with visual data, addressing the persistent challenge of occlusion during robot-object contact. His earlier "Tactile SLAM" framework demonstrated simultaneous shape and pose inference through planar pushing, drawing elegant parallels between classical SLAM and contact-rich manipulation. Beyond manipulation, Suresh has also contributed to underwater robotics, developing refraction-corrected stereo SLAM for AUV localization (2019, 26 citations). Collectively, his research reflects a commitment to robust, perception-driven robotics across diverse and challenging real-world domains.
Research Focus
Key Achievements
Top Papers
- 1
- 2ShapeMap 3-D: Efficient shape mapping through dense touch and vision50 citations · 2022
- 3Through-Water Stereo SLAM With Refraction Correction for AUV Localization26 citations · 2019
- 4Tactile SLAM: Real-time inference of shape and pose from planar pushing7 citations · 2021
- 5Efficient shape mapping through dense touch and vision.3 citations · 2021