Papers
3
Total Citations
9
H-Index
2
About
Kumaraditya Gupta is a researcher at the forefront of 3D scene understanding and vision-language navigation for robotics. His work focuses on bridging the gap between raw sensor data and the hierarchical, semantic representations that intelligent agents need for planning and exploration. Gupta’s key contribution is the development of **QueSTMaps** (Queryable Semantic Topological Maps), a framework that enables robots to build and query maps that capture not just objects, but also topological regions like rooms and floors. This approach, which has garnered over 5 citations since its 2024 publication, directly addresses a critical limitation in existing 3D segmentation methods. He further advanced the field with **O3D-SIM** (Open-set 3D Semantic Instance Maps), a system that allows robots to form open-vocabulary, instance-level semantic maps for language-guided navigation. Building on prior work in instance-level mapping, Gupta’s research is pioneering the creation of truly queryable, human-like spatial understanding in autonomous systems, enabling more natural human-robot interaction and robust navigation in complex, multi-floor environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3