Research map / Detection and segmentation / Weakly supervised detection and segmentation
PsyNet: Self-Supervised Approach to Object Localization Using Point Symmetric Transformation
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 spatial attention map (channel pooling) instead of average Multi-CAM heat map (ablation) | PsyNet rotation-only · CUB-200-2011 | GT-Known Loc: 60.82 -> 80.51 | Helped |
| add PST to rotation self-supervision (ablation) | PsyNet · CUB-200-2011 | GT-Known Loc: 80.51 -> 83.78 | Helped |
| add PST to rotation self-supervision (ablation) | PsyNet · FGVC-Aircraft | GT-Known Loc: 87.67 -> 96.61 | Helped |
| add PST to rotation self-supervision (ablation) | PsyNet · Stanford Cars | GT-Known Loc: 89.21 -> 95.59 | Helped |
| add scaling transformation to PST+rotation (ablation) | PsyNet · CUB-200-2011 | GT-Known Loc: 83.78 -> 58.66 | Hurt |
| use PST translation 1.0 (maximal erasing) instead of 1.8 (ablation) | PsyNet PST · CUB-200-2011 | GT-Known Loc: 32.51 -> 72.06 | Helped |
| randomly sample PST translation instead of fixed 1.0 (ablation) | PST alone · CUB-200-2011 | GT-Known Loc: 72.06 -> 75.34 | Mixed |
| replace CAM with spatial-attention heat map (M) for WSOL | CAM baseline · CUB-200-2011 WSOL | GT-Known Loc: 51.09 -> 80.32 | Helped |
| add PST geometric augmentation to PsyNet(M) (ablation) | PsyNet(M) · CUB-200-2011 WSOL | Top-1 Loc: 57.97 -> 59.37 | Helped |
| add PST geometric augmentation to PsyNet(M) (ablation) | PsyNet(M) · CUB-200-2011 WSOL | GT-Known Loc: 80.32 -> 77.39 | Hurt |
| use SE-ResNet50 backbone instead of VGG16 | PsyNet · CUB-200-2011 | GT-Known Loc: 83.78 -> 85.10 | Helped |
Closest papers by meaning
- Self-Supervised Object Localization with Joint Graph Partition (AAAI 2022 · similarity 0.91)
- Bridging the Gap Between Classification and Localization for Weakly Supervised Object Localization (CVPR 2022 · similarity 0.88)
- Unveiling the Potential of Structure Preserving for Weakly Supervised Object Localization (CVPR 2021 · similarity 0.88)
- Learning to Localize Objects with Noisy Labeled Instances (AAAI 2019 · similarity 0.87)
- Rethinking the Route Towards Weakly Supervised Object Localization (CVPR 2020 · similarity 0.87)
- Leveraging Imagery Data with Spatial Point Prior for Weakly Semi-supervised 3D Object Detection (AAAI 2024 · similarity 0.85)
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.