Brojeshwar Bhowmick
Papers
13
Total Citations
184
H-Index
7
About
Brojeshwar Bhowmick is a researcher whose work spans computer vision, embodied AI, and autonomous robotics, with particular focus on human-robot interaction, robot navigation, and scene understanding. His early influential contribution — "Person Identification using Skeleton Information from Kinect" (2013, 95 citations) — established a robust methodology for recognizing individuals through gait analysis using depth sensor data, addressing a critical need in proliferating human-robot systems. This work remains his most widely cited and signals his enduring interest in making robots perceptually aware of the humans around them. More recently, Bhowmick has made significant strides in embodied AI, tackling nuanced challenges such as resolving language ambiguity during robotic task execution ("DoRO," 19 citations; "Talk-to-Resolve," 16 citations) and advancing object-goal navigation through semantic mapping, graph convolutional networks, and reinforcement learning. His "Teledrive" telepresence system demonstrates a practical application of these capabilities, enabling robots to autonomously navigate unknown environments to support remote caregiving. Across his body of work, Bhowmick consistently bridges perception, natural language understanding, and autonomous decision-making — making meaningful contributions to the emerging field of socially intelligent, navigation-capable robots operating in real-world settings.
Research Focus
Key Achievements
Top Papers
- 1Person Identification using Skeleton Information from Kinect95 citations · 2013
- 2DoRO: Disambiguation of Referred Object for Embodied Agents19 citations · 2022
- 3
- 4Sequence-Agnostic Multi-Object Navigation9 citations · 2023
- 5Object Goal Navigation using Data Regularized Q-Learning8 citations · 2022
- 6Learning to Prevent Monocular SLAM Failure using Reinforcement Learning8 citations · 2018
- 7
- 8
- 9
- 10Teledrive: An Embodied AI Based Telepresence System3 citations · 2024