Parv Maheshwari
Papers
2
Total Citations
48
H-Index
2
About
Parv Maheshwari is a robotics and autonomous systems researcher whose work sits at the intersection of off-road navigation, self-supervised learning, and risk-aware planning. He is perhaps best known for his contributions to the TartanDrive dataset series — large-scale, multi-modal off-road driving datasets that have become foundational resources for the autonomous driving research community. The 2024 release of TartanDrive 2.0, which expanded upon the original dataset with richer modalities and improved infrastructure, has already garnered 25 citations, reflecting its rapid uptake among researchers pushing the boundaries of self-supervised learning in challenging terrains. Beyond data collection, Maheshwari has made meaningful algorithmic contributions, particularly in the domain of costmap learning. His 2023 paper on learning risk-aware costmaps via inverse reinforcement learning addresses a longstanding pain point in off-road autonomy — the labor-intensive design of reliable costmaps — by instead learning them directly from expert driving demonstrations. With 23 citations, this work has resonated strongly within the navigation community. Together, these contributions position Maheshwari as an emerging voice in robust, data-driven approaches to autonomous off-road driving, with research that bridges large-scale dataset curation and principled, safety-conscious machine learning.
Research Focus
Key Achievements
Top Papers
- 1
- 2