Sparsh Garg
Papers
1
Total Citations
11
H-Index
1
About
Sparsh Garg is a robotics researcher whose work focuses on bridging the critical gap between simulation and reality for robotic manipulation. His primary research areas include sim-to-real transfer, computer vision, and policy learning for dexterous manipulation. Garg’s most notable contribution is **SplatSim**, a novel framework introduced in his highly cited 2025 paper that leverages Gaussian Splatting to achieve zero-shot sim-to-real transfer of RGB-based manipulation policies. This work directly addresses the persistent challenge of domain shift between synthetic and real-world visual data, enabling policies trained entirely in simulation to succeed in real environments without any fine-tuning. With 11 citations in its first year, SplatSim has quickly gained attention for its practical impact on deploying vision-based robotic systems. Garg’s research is particularly significant for advancing the use of photorealistic rendering in robotics, offering a scalable path from simulation to real-world deployment. His work stands out for its elegant solution to a long-standing bottleneck in robot learning, making him a rising figure in the field of embodied AI and manipulation.
Research Focus
Key Achievements
Top Papers
- 1