Research map / Detection and segmentation / Open-set and few-shot detection
See More for Scene: Pairwise Consistency Learning for Scene Classification
Gongwei Chen, Xinhang Song, Bohan Wang, Shuqiang Jiang
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 pairwise consistency learning | ResNet34 · Places365-small | Top-1 accuracy: 48.52 -> 50.72 | Improved |
| add pairwise consistency learning | ResNet34 · Places365 | Top-1 accuracy: 54.84 -> 56.23 | Improved |
| add pairwise consistency learning | DenseNet121 · Places365-small | Top-1 accuracy: 49.48 -> 51.55 | Improved |
| add pairwise consistency learning | ResNet18 · ImageNet | Top-1 accuracy: 69.93 -> 69.57 | Hurt |
| replace HaS erasing with pairwise consistency | ResNet18 · Places365-small | Top-1 accuracy: 49.25 -> 49.70 | Improved |
| replace ADL erasing with pairwise consistency | ResNet34 · Places365-small | Top-1 accuracy: 48.68 -> 50.72 | Improved |
| insert CAM module at stage3 instead of stage4 (ablation) | ResNet34 · Places365-small | Top-1 accuracy: 49.77 -> 50.72 | Improved |
| insert CAM module at stage3 instead of stage4 (ablation) | ResNet18 · Places365-small | Top-1 accuracy: 49.62 -> 49.70 | No clear effect |
| use category-level pairwise consistency instead of attention mining loss (ablation) | ResNet34 · Places365-small | Top-1 accuracy: 48.89 -> 49.92 | Improved |
| use instance-level instead of category-level consistency (ablation) | ResNet34 · Places365-small | Top-1 accuracy: 49.92 -> 50.36 | Improved |
| combine CPC and IPC (ablation) | ResNet34 · Places365-small | Top-1 accuracy: 50.36 -> 50.72 | Improved |
| simple instead of complex training scheme | ResNet18 · ImageNet | Top-1 accuracy: 69.93 -> 68.80 | Hurt |
Closest papers by meaning
- Object-level Scene Context Prediction (TPAMI 2021 · similarity 0.86)
- Classification by Attention: Scene Graph Classification with Prior Knowledge (AAAI 2021 · similarity 0.86)
- Disentangling Instance and Scene Contexts for 3D Semantic Scene Completion (ICCV 2025 · similarity 0.86)
- Tell Me Where I Am: Object-Level Scene Context Prediction (CVPR 2019 · similarity 0.86)
- Scene Consistency Representation Learning for Video Scene Segmentation (CVPR 2022 · similarity 0.86)
- SCENIR: Visual Semantic Clarity through Unsupervised Scene Graph Retrieval (ICML 2025 · similarity 0.84)
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.
