Robotic Embedded Systems Laboratory @USC
Los Angeles, CA
Joined June 2009
- Tweets109
- Following47
- Followers325
- Likes23
USC RESL retweeted
๐ค Ever wondered if you can train end-to-end humanoid policies entirely within a differentiable simulator?
๐ฅ Introducing ๐๐๐ป๐ฑ๐น๐ฒ๐ฑ ๐๐ผ๐ป๐๐ฎ๐ฐ๐ ๐๐ฟ๐ฎ๐ฑ๐ถ๐ฒ๐ป๐๐ (๐๐๐)
๐: bundledcontactgradients.githโฆ
๐: arxiv.org/abs/2609.30951
๐งต๐
Learning-based systems still require large amounts of data for effective generalization. Our CoRL 2025 paper, D-CODA, expands data without simulators or environment interactions by generating novel view-consistent wrist-camera images and action labels for bimanual robots.
๐งต๐
Project website: dcodaaug.github.io/D-CODA/
Paper: arxiv.org/pdf/2505.04860
Authors: @arthur801031, @jasoon_chen, @gauravsukhatme, @daniel_t_seita
End thread๐งต
The original states are augmented every k๏ฟผ time steps, resulting in out-of-distribution states. The non-augmented states remain in-distribution and guide the behavior cloning policies to complete the tasks.
Excited to announce our #RSS2025 workshop:
Resource Constrained Robotics
Discover how robotics thrives under constraints โ from limited compute to smarter algorithms enabling real-time, low-power autonomy.
๐
June 21 | ๐ Los Angeles, CA
๐ sites.google.com/view/rc4robโฆ
Tasks requiring two-hand coordination and fine-grained manipulation remain challenging for current robotic systems. Our CoRL 2024 paper proposes a sample-efficient, language-conditioned, voxel-based method that utilizes Vision Language Models to address these challenges.
๐งต๐
Blog post: rasc.usc.edu/blog/voxact-b/
Project website: voxact-b.github.io/
Paper: arxiv.org/abs/2407.04152
Code: github.com/VoxAct-B/voxactb
Authors: @arthur801031, @SichengHe12345, @daniel_t_seita, @gauravsukhatme
End thread๐งต