DexCast starts from monocular RGB-D human videos of manipulation.
DexCast learns real-world dexterous control from human videos without teleoperation
The framework reconstructs 3D hand-object interactions from monocular RGB-D footage and conditions a diffusion policy on learned joint-coordination latents, achieving 87.5% average success across six tasks.
Top University
Zihan Gao · Yuhong Zhang · Mengzhe Wang · Kaisheng Liu · Shengpeng Li · Siyao Xiao · +4 more
Tsinghua University · TsingTec · Shenzhen University · Nanyang Technological University · Georgia Institute of Technology
Research Digest··2 min read
Thread:Robot Manipulation Policies
Gao, Zhang, Wang and colleagues present DexCast, a framework that transfers dexterous manipulation skills to a five-fingered robot hand using only monocular human RGB-D videos.
Why this paper
From Tsinghua University and 4 others · Part of Robot Manipulation Policies, now 19 papers
In one line
DexCast learns dexterous robotic hand control from monocular human videos without teleoperation, achieving 87.5% success on real-world tasks.
What we could check
- ·No code link found
- ·No weights link found
- ·No dataset link found
- ·No compute details found
- ✓Limitations stated by the authors (2 noted)
- ✓Reports numbers on named benchmarks
Observed from the paper text and links we have. Absence here means we did not find it, not that it does not exist.
§