Simar Kareer

Papers

Total Citations

H-Index

About

Simar Kareer is an emerging robotics researcher whose work sits at the intersection of embodied AI, robot locomotion, and imitation learning. Kareer's research addresses some of the most pressing challenges in deploying autonomous robots in real-world, unstructured environments — from navigating cluttered apartments to acquiring dexterous manipulation skills at scale. His most recognized contribution, **ViNL: Visual Navigation and Locomotion Over Obstacles** (2023, 28 citations), introduced a framework enabling quadrupedal robots to intelligently step over small obstacles during navigation — a deceptively difficult problem that bridges high-level path planning with low-level motor control. This work represents a meaningful step toward robots that can operate gracefully in human living spaces. Kareer's more recent **EgoMimic** framework tackles a fundamental bottleneck in robot learning: the scarcity of diverse demonstration data. By leveraging egocentric human video paired with 3D hand tracking, EgoMimic enables manipulation policies to be trained from human embodiment data, dramatically scaling imitation learning without requiring costly robot demonstrations. Already accumulating notable citations across its versions, this work signals growing community interest in human-centric data pipelines for robotics. Kareer's portfolio marks him as a researcher to watch in scalable robot learning.