Research map / Vision-language and multimodal / CLIP and zero-shot
Adapting Vision-Language Models from Iconic to Inclusive for Multi-label Recognition Without Labels
Cheng Chen, Jingyu Zhou, Yifan Zhao, Jia Li
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 multi-sampling response estimator (ablation) | cutting stage · VOC2012 | mAP: 85.1 -> 88.3 | Improved |
| add multi-object blend adaptation (sewing) (ablation) | Ours · VOC2012 | mAP: 88.6 -> 91.4 | Improved |
| add confidence correction (ablation) | Ours · VOC2012 | mAP: 91.1 -> 91.4 | Improved |
| use order-preserving soft correction instead of hard correction (ablation) | Ours · MS-COCO | mAP: 71.7 -> 72.5 | Improved |
| use soft square-root correction instead of hard (ablation) | Ours · VOC2007 | mAP: 91.2 -> 92.6 | Improved |
| use soft square-root correction instead of no correction (ablation) | Ours · VOC2007 | mAP: 92.2 -> 92.6 | Improved |
| use soft square-root correction instead of linear (ablation) | Ours · VOC2007 | mAP: 92.2 -> 92.6 | Improved |
| increase sampling iterations M from 1 to 20 (ResNet-101) (ablation) | Ours · VOC2007 | mAP: 89.5 -> 92.5 | Improved |
| increase sampling iterations M from 20 to 50 (ViT-L/14) (ablation) | Ours · VOC2007 | mAP: 92.2 -> 89.9 | Hurt |
| increase sampling iterations M from 20 to 50 (ResNet-101) (ablation) | Ours · VOC2007 | mAP: 92.5 -> 92.7 | No clear effect |
| increase beta from 1 to 3 (ablation) | Ours · VOC2007 | mAP: 91.8 -> 91.1 | Hurt |
| hard label correction instead of no correction (ablation) | Ours · MS-COCO train set | mAP: 67.5 -> 64.1 | Hurt |
Closest papers by meaning
- Language-Driven Cross-Modal Classifier for Zero-Shot Multi-Label Image Recognition (ICML 2024 · similarity 0.90)
- SPARC: Score Prompting and Adaptive Fusion for Zero-Shot Multi-Label Recognition in Vision-Language Models (CVPR 2025 · similarity 0.89)
- Vision-Language Models are Strong Noisy Label Detectors (NeurIPS 2024 · similarity 0.87)
- Epsilon: Exploring Comprehensive Visual-Semantic Projection for Multi-Label Zero-Shot Learning (AAAI 2025 · similarity 0.87)
- Recover and Match: Open-Vocabulary Multi-Label Recognition through Knowledge-Constrained Optimal Transport (CVPR 2025 · similarity 0.87)
- Efficient and Context-Aware Label Propagation for Zero-/Few-Shot Training-Free Adaptation of Vision-Language Model (ICLR 2025 · similarity 0.87)
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.
