Snehasis Banerjee

Papers

Total Citations

H-Index

About

Snehasis Banerjee is a leading researcher in embodied AI and robotic navigation, with a focus on enabling assistive robots to operate intelligently in dynamic, real-world environments. His key contributions center on multi-object navigation and task decomposition, addressing fundamental challenges in how robots locate and interact with multiple objects in unstructured settings. Banerjee’s work on the Multi-Object Navigation (MultiON) task, as detailed in his highly cited 2023 paper, introduced a sequence-agnostic approach that allows robots to efficiently localize instances of multiple object classes—a critical capability for home and factory assistants. More recently, his 2025 work on AdaptBot pioneers the integration of Large Language Models (LLMs) with knowledge graphs and human input, enabling robots to decompose novel tasks into actionable steps without extensive retraining. This generic-to-specific framework represents a significant leap in zero-shot learning for embodied agents. With his papers accumulating citations and his research bridging theoretical AI with practical robotics, Banerjee is shaping the future of adaptive, human-aware robotic systems that can learn and refine their knowledge on the fly.