Research map / Detection and segmentation / Open-set and few-shot detection
Unsupervised Object-Centric Learning From Multiple Unspecified Viewpoints
Jinyang Yuan, Tonglin Chen, Zhimeng Shen, Bin Li, Xiangyang Xue
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 |
|---|---|---|---|
| add explicit shadow modeling (ablation) | OCLOC · CLEVR & SHOP Test 1 | ARI-A: 0.395 -> 0.734 | Improved |
| add explicit shadow modeling (ablation) | OCLOC · GSO & ShapeNet Test 1 | ARI-A: 0.429 -> 0.831 | Improved |
| add explicit shadow modeling (ablation) | OCLOC · GSO & ShapeNet Test 1 | IoU: 0.514 -> 0.707 | Improved |
| add explicit shadow modeling (ablation) | OCLOC · GSO & ShapeNet Test 1 | OCA: 0.000 -> 0.715 | Improved |
| add explicit shadow modeling (ablation) | OCLOC · CLEVR & SHOP Test 1 | ARI-O: 0.903 -> 0.859 | Hurt |
| add explicit shadow modeling (ablation) | OCLOC · CLEVR & SHOP Test 2 | ARI-A: 0.265 -> 0.545 | Improved |
| order viewpoints as video for SAVi instead of unordered | SAVi · CLEVR & SHOP Test 1 | ARI-O: 0.046 -> 0.715 | Improved |
| use OCLOC without viewpoint annotations instead of annotated MulMON | OCLOC · CLEVR & SHOP Test 1 | ARI-A: 0.473 -> 0.734 | Improved |
| use OCLOC instead of MulMON | OCLOC · CLEVR & SHOP Test 1 | OCA: 0.227 -> 0.205 | Hurt |
| use OCLOC instead of SIMONe | OCLOC · GSO & ShapeNet Test 1 | ARI-A: 0.404 -> 0.831 | Improved |
| test with more objects (7-10, K=11) than Test 1 | OCLOC · CLEVR & SHOP | ARI-A: 0.734 -> 0.545 | Hurt |
| test with more objects (7-10, K=11) than Test 1 | OCLOC · GSO & ShapeNet | ARI-A: 0.831 -> 0.708 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- SAVi — SAVi++: Towards End-to-End Object-Centric Learning from Real-World Videos (NeurIPS 2022) · better on 4 of 4 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Unsupervised Learning of Compositional Scene Representations from Multiple Unspecified Viewpoints (AAAI 2022 · similarity 0.98)
- Learning Object-Centric Representations of Multi-Object Scenes from Multiple Views (NeurIPS 2020 · similarity 0.90)
- Generative Modeling of Infinite Occluded Objects for Compositional Scene Representation (ICML 2019 · similarity 0.90)
- Compositional Scene Understanding through Inverse Generative Modeling (ICML 2025 · similarity 0.89)
- Improving Viewpoint-Independent Object-Centric Representations through Active Viewpoint Selection (NeurIPS 2024 · similarity 0.88)
- Learning to Compose: Improving Object Centric Learning by Injecting Compositionality (ICLR 2024 · similarity 0.86)
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.
