Suryansh Kumar
Papers
7
Total Citations
132
H-Index
4
About
Suryansh Kumar is a researcher specializing in computer vision, robotics, and 3D scene understanding, with a particular focus on active perception, neural rendering, and robotic reconstruction. His work bridges the gap between classical robotic vision and modern deep learning techniques, making significant contributions to how autonomous systems perceive and reconstruct their environments. Kumar's most impactful contribution, "Uncertainty Guided Policy for Active Robotic 3D Reconstruction Using Neural Radiance Fields" (2022, 83 citations), introduced a principled approach to view planning for robotic 3D reconstruction by leveraging Neural Radiance Fields (NeRF) and uncertainty estimation — a notable departure from conventional methods. His early work in 2014 addressed the challenging problem of small obstacle detection using Markov Random Fields and active robot guidance, laying foundational groundwork in indoor robotic perception. More recently, his research has expanded into dense semantic SLAM with neural implicit representations and mobile robotic multi-view photometric stereo, demonstrating a sustained commitment to advancing fine-grained 3D acquisition. Across his career, Kumar has consistently pushed the boundaries of how robots actively gather visual information to build richer, more accurate world models — work that remains highly relevant as embodied AI and autonomous systems continue to evolve.
Research Focus
Key Achievements
Top Papers
- 1
- 2Markov Random Field based small obstacle discovery over images25 citations · 2014
- 3Small Object Discovery and Recognition Using Actively Guided Robot9 citations · 2014
- 4Neural Implicit Dense Semantic SLAM6 citations · 2023
- 5
- 6Mobile robotic multi-view photometric stereo3 citations · 2025
- 7CRF Based Frontier Detection using Monocular Camera2 citations · 2014