Postdoc, University of California, Berkeley
2 papers at NeurIPS 2025
We propose a simple robot learning recipe leveraging privileged information to train active perception policies on real robots.
