Papers
3
Total Citations
8
H-Index
2
About
Yash Mehan is a researcher at the forefront of 3D scene understanding and robotic perception, with a focus on enabling machines to navigate and interpret complex environments hierarchically. His work centers on developing semantic and topological mapping techniques that bridge the gap between low-level sensor data and high-level spatial reasoning. In his highly cited 2024 paper, *QueSTMaps: Queryable Semantic Topological Maps for 3D Scene Understanding*, Mehan introduces a novel framework that allows robots to segment scenes into meaningful topological regions—such as rooms and floors—rather than relying solely on object-level segmentation. This approach enhances planning and navigation by providing a queryable, hierarchical representation of space. Earlier, in his 2023 work *Hierarchical Unsupervised Topological SLAM*, he pioneered an unsupervised method for topological clustering that improves loop detection and closure in simultaneous localization and mapping (SLAM). Though early in his career, Mehan’s contributions are already shaping how robots build and use semantic maps, with growing citation impact that underscores the relevance of his research to autonomous systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3Hierarchical Unsupervised Topological SLAM1 citations · 2023