Research map / Vision-language and multimodal / CLIP and zero-shot
Proximity-CLIP: Text-Guided Semantic Proximity Learning for Zero-Shot Anomaly Detection
Manwen Yang, Leqian Ding, Yu Guo, Fei Wang
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 stage-one semantic calibration (ablation) | Proximity-CLIP · Medical | Pixel AUROC: 55.5 -> 84.4 | Improved |
| add stage-one semantic calibration (SC) to AQM (ablation) | Proximity-CLIP · Industrial | Image AUROC: 86.2 -> 70.4 | Improved |
| add Norm-Preserving mechanism for both text and image (ablation) | Proximity-CLIP · Medical | Image AUROC: drop of 4.4% | Improved |
| add Norm-Preserving mechanism for both text and image (ablation) | Proximity-CLIP · Industrial | Image AUROC: drop of 4.2% | Improved |
| replace multi-scale pooling with AQM cross-attention querying (ablation) | Proximity-CLIP · Industrial | Pixel AUROC: +1.3% | Improved |
| replace multi-scale pooling with AQM cross-attention querying (ablation) | Proximity-CLIP · Medical | Image AUROC: +2.2% | Improved |
| bound margin lambda with sigmoid and affine scalars instead of unconstrained scalar (ablation) | Proximity-CLIP · text adaptation stage | proximity loss convergence | Improved |
| restrict alpha to 0.35 or lower (upper bound 0.65 or lower) (ablation) | Proximity-CLIP · six datasets average | detection performance | Hurt |
| vary alpha at or above 0.5 with beta 0.3 (ablation) | Proximity-CLIP · six datasets average | detection performance | No clear effect |
| raise lower bound beta toward 1.0 (ablation) | Proximity-CLIP · six datasets average | detection performance | Hurt |
| use Proximity-CLIP instead of AA-CLIP | AA-CLIP · Seven datasets average | Pixel AUROC: 95.5 -> 95.7 | Improved |
| use Proximity-CLIP instead of AA-CLIP | AA-CLIP · Seven datasets average | Image AUROC: 80.6 -> 81.0 | 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.
- AA-CLIP — AA-CLIP: Enhancing Zero-Shot Anomaly Detection via Anomaly-Aware CLIP (CVPR 2025) · better on 11 of 17 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
- VisualAD: Language-Free Zero-Shot Anomaly Detection via Vision Transformer (CVPR 2026 · similarity 0.91)
- Defect-aware Hybrid Prompt Optimization for Zero-Shot Multi-type Anomaly Detection and Segmentation (ECCV 2026 · similarity 0.91)
- AG-VAS: Anchor-Guided Zero-Shot Visual Anomaly Segmentation with Large Multimodal Models (CVPR 2026 · similarity 0.90)
- AA-CLIP: Enhancing Zero-Shot Anomaly Detection via Anomaly-Aware CLIP (CVPR 2025 · similarity 0.89)
- Towards Zero-Shot Anomaly Detection and Reasoning with Multimodal Large Language Models (CVPR 2025 · similarity 0.88)
- Aligning and Prompting Anything for Zero-Shot Generalized Anomaly Detection (AAAI 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.
