Research map / Label-efficient and robust learning / Recognition and long-tail
Contextual Debiasing for Visual Recognition With Causal Mechanisms
Ruyang Liu, Hao Liu, Ge Li, Haodi Hou, TingHao Yu, Tao Yang
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 |
|---|---|---|---|
| Apply CCD to ResNet101 at 576 × 576 | ResNet101 · MS-COCO | mAP: 85.2 -> 85.3 | Improved |
| Apply CCD to Swin-L(22k) | Swin-L(22k) · MS-COCO | mAP: 89.5 -> 90.3 | Improved |
| Apply CCD to ResNet101 | ResNet101 · PASCAL-VOC 2007 | mAP: 96.0 -> 95.8 | Hurt |
| Apply CCD to ResNet101 | ResNet101 · PASCAL-VOC 2012 | mAP: 95.5 -> 96.1 | Improved |
| Apply CCD to ResNet101 | ResNet101 · NUS-WIDE | mAP: 63.9 -> 65.1 | Improved |
| Add the intervention module (ablation) | ResNet101 multi-label classifier · MS-COCO | mAP: 81.5 -> 82.5 | Improved |
| Add the counterfactual module (ablation) | ResNet101 multi-label classifier · MS-COCO | mAP: 81.5 -> 82.8 | Improved |
| Add both intervention and counterfactual modules (ablation) | ResNet101 multi-label classifier · MS-COCO | mAP: 81.5 -> 84.0 | Improved |
| Add the intervention module (ablation) | VIT multi-label classifier · MS-COCO | mAP: 80.9 -> 84.6 | Improved |
| Add the counterfactual module (ablation) | VIT multi-label classifier · MS-COCO | mAP: 80.9 -> 84.2 | Improved |
| Add both intervention and counterfactual modules (ablation) | VIT multi-label classifier · MS-COCO | mAP: 80.9 -> 85.2 | Improved |
| Replace simple context mapping M with W · M (ablation) | CCD-R101 · MS-COCO | mAP: 84.0 -> 83.9 | Hurt |
Closest papers by meaning
- In Pursuit of Causal Label Correlations for Multi-label Image Recognition (NeurIPS 2024 · similarity 0.89)
- ContextShift: A Controlled Benchmark for Context Dependence in Object Detection (NeurIPS 2026 · similarity 0.87)
- Counterfactual Reasoning for Multi-Label Image Classification via Patching-Based Training (ICML 2024 · similarity 0.86)
- Causal-HalBench: Uncovering LVLMs Object Hallucinations Through Causal Intervention (AAAI 2026 · similarity 0.86)
- Representation-Level Counterfactual Calibration for Debiased Zero-Shot Recognition (NeurIPS 2025 · similarity 0.86)
- Classifier-to-Bias: Toward Unsupervised Automatic Bias Detection for Visual Classifiers (CVPR 2025 · 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.
