Research map / Detection and segmentation / Open-set and few-shot detection
Unsupervised Discovery of Parts, Structure, and Dynamics
Zhenjia Xu*, Zhijian Liu*, Chen Sun, Kevin Murphy, William T. Freeman, Joshua B. Tenenbaum, Jiajun Wu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use PSD's layered representation and structural descriptor instead of NEM/R-NEM | Object segmentation on shapes and digits · Shapes and digits | IoU: 0.935 vs 0.540; 0.816 vs 0.559; 0.905 vs 0.583; 0.750 vs 0.470; 0.742 vs 0.416; 0.739 vs 0.339; 0.739 vs 0.512; 0.472 vs | Improved |
| Use PSD's layered representation and structural descriptor instead of NEM/R-NEM | Part segmentation · Human exercise | IoU: 0.697 vs 0.321; 0.574 vs 0.347; 0.391 vs 0.220; 0.374 vs 0.294; 0.336 vs 0.228; 0.474 vs 0.276 | Improved |
| Give R-NEM 20 sequential frames rather than 2 repeated frames | R-NEM · Shapes | Overall IoU: 0.558 -> 0.833 | Improved |
| Give R-NEM 20 sequential frames rather than 1 frame | R-NEM · Shapes | Overall IoU: 0.501 -> 0.833 | Improved |
Closest papers by meaning
- PARTS: Unsupervised Segmentation With Slots, Attention and Independence Maximization (ICCV 2021 · similarity 0.88)
- Learning Features by Watching Objects Move (CVPR 2017 · similarity 0.88)
- Unsupervised learning of object structure and dynamics from videos (NeurIPS 2019 · similarity 0.87)
- Unsupervised Part Discovery from Contrastive Reconstruction (NeurIPS 2021 · similarity 0.87)
- Discovering Objects That Can Move (CVPR 2022 · similarity 0.87)
- HEIR: Learning Graph-Based Motion Hierarchies (NeurIPS 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
