PhD Student in Visual Computing Lab in Seoul National University | 3D Computer Vision, Human-Object Interaction

Joined June 2022
2 Photos and videos
Pinned Post
Introducing ZeroDex: Zero-Shot Long-Horizon Dexterous Manipulation via Multi-View 3D-Grounded VLM Reasoning VLMs know what to do. ZeroDex grounds that reasoning in multi-view 3D, enabling robots to perform dexterous actions. No task-specific demos. No policy fine-tuning.
2
20
102
11,131
Introducing HRDexDB โ€” the first dataset of paired humanโ€“robot dexterous grasping sequences over the same objects. Diverse hand embodiments, 100 objects, and accurate 3D annotations. ๐ŸŒ Project: snuvclab.github.io/HRDexDB/ ๐Ÿค— Dataset: huggingface.co/datasets/HRDeโ€ฆ
5
15
63
8,192
Introducing ZeroDex: Zero-Shot Long-Horizon Dexterous Manipulation via Multi-View 3D-Grounded VLM Reasoning VLMs know what to do. ZeroDex grounds that reasoning in multi-view 3D, enabling robots to perform dexterous actions. No task-specific demos. No policy fine-tuning.
2
20
102
11,131
For long-horizon tasks, ZeroDex closes the loop with task verification and retry. We demonstrate zero-shot dexterous manipulation across unseen objects, tool-use tasks, and long-horizon real-robot scenarios. ๐ŸŒ Project: jlogkim.github.io/zerodex/ ๐Ÿ“„ Paper: arxiv.org/abs/2606.19340
1
1
5
397
Jisoo Kim reposted
We introduce ๐Ÿ–๏ธ๐ŸŒ ๐ƒ๐ž๐ฑ๐ญ๐ž๐ซ๐จ๐ฎ๐ฌ ๐–๐จ๐ซ๐ฅ๐ ๐Œ๐จ๐๐ž๐ฅ๐ฌ (๐ƒ๐–๐Œ) โ€” a scene-action-conditioned video diffusion model that simulates human manipulation in static 3D scenes from egocentric hand motions. ๐Ÿ“„ Paper: arxiv.org/abs/2512.17907 ๐ŸŒ Project Page: snuvclab.github.io/dwm/
6
43
274
30,177
The full version of the ParaHome dataset is now available! ๐ŸŽ‰ Check out our webpage: github.com/snuvclab/ParaHome - 486 min, 207 sequences, 38 subjects - 3D parameterized 22 objects - Text annot. for every action - Full 3D mocap for body and hand movements - Sequential HOI scenarios
14
58
9,069