Research map / Detection and segmentation / Weakly supervised detection and segmentation
Cross-token Guidance Transformer for Weakly Supervised Object Localization
Zhiwei Chen, Yiran Nie, Ruize Han, Qinqin Zhou
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 Attention Regulation Module to vanilla (ablation) | vanilla DeiT-S location-token transformer · ILSVRC val | GT-known Loc: 69.10 -> 70.91 | Improved |
| add Filter Regulation Module to vanilla (ablation) | vanilla DeiT-S location-token transformer · ILSVRC val | Top-1 Loc: 56.86 -> 58.22 | Improved |
| add FRM on top of ARM (ablation) | ARM-only model · ILSVRC val | Top-1 Loc: 58.57 -> 60.35 | Improved |
| add ARM on top of FRM (ablation) | FRM-only model · ILSVRC val | GT-known Loc: 70.58 -> 73.33 | Improved |
| add ARM to vanilla (ablation) | vanilla · CUB-200-2011 | Top-1 Loc: 76.25 -> 79.43 | Improved |
| add FRM to vanilla (ablation) | vanilla · CUB-200-2011 | Top-1 Loc: 76.25 -> 78.36 | Improved |
| add both ARM and FRM (ablation) | vanilla · CUB-200-2011 | GT-known Loc: 91.77 -> 98.68 | Improved |
| add both ARM and FRM (ablation) | vanilla · ILSVRC val | Top-1 Loc: 56.86 -> 60.35 | Improved |
| use lambda 0.45 instead of 0.40/0.55 (ablation) | CGTR · CUB-200-2011 | Top-1 Loc: 80.81 -> 81.80 | Mixed |
| vary lambda 0.40 to 0.55 (ablation) | CGTR · ILSVRC val | Top-1 Loc: 60.28 -> 60.35 | No clear effect |
| use larger pooling kernel k in FRM (ablation) | CGTR · ILSVRC val | Loc accuracy | Hurt |
| set xi to 0.37 (ablation) | CGTR · ILSVRC val | Loc accuracy | Improved |
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.
- TS-CAM — TS-CAM: Token Semantic Coupled Attention Map for Weakly Supervised Object Localization (ICCV 2021) · better on 5 of 5 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
- Weakly Supervised Object Localization via Transformer with Implicit Spatial Calibration (ECCV 2022 · similarity 0.94)
- Spatial-Aware Token for Weakly Supervised Object Localization (ICCV 2023 · similarity 0.92)
- CLIP-Driven Transformer for Weakly Supervised Object Localization (TPAMI 2025 · similarity 0.92)
- Category-aware Allocation Transformer for Weakly Supervised Object Localization (ICCV 2023 · similarity 0.91)
- Evaluating Weakly Supervised Object Localization Methods Right (CVPR 2020 · similarity 0.91)
- Evaluation for Weakly Supervised Object Localization: Protocol, Metrics, and Datasets (TPAMI 2022 · similarity 0.91)
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.
