Research map / Vision-language and multimodal / CLIP and zero-shot
From Attraction to Equilibrium: Physics-Inspired Semantic Gravitons for Zero-Shot Anomaly Detection
Yuwen Pan, Yuan Wang, Shaohui Li, Zhi Li, Yu Liu, You He
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 Graviton Interaction Network and potential regularization (ablation) | SGNet · MVTec-AD image-level | AUROC: 91.1 -> 93.5 | Improved |
| add Graviton Interaction Network (ablation) | SGNet · MVTec-AD image-level | AUROC: 91.1 -> 91.8 | Improved |
| add graviton potential regularization (attraction and equilibrium) on top of GIN (ablation) | SGNet · MVTec-AD image-level | AUROC: 91.8 -> 93.5 | Improved |
| add graviton potential regularization on top of GIN (ablation) | SGNet · VisA image-level | AUROC: 85.1 -> 85.9 | Improved |
| increase graviton number from 20 to 30 (ablation) | SGNet · MVTec-AD image-level | AUROC: 93.5 -> 93.1 | Hurt |
| reduce graviton number from 20 to 10 (ablation) | SGNet · MVTec-AD image-level | AUROC: 93.5 -> 92.1 | Hurt |
| vary lambda_grav and lambda_att (ablation) | SGNet · MVTec-AD and VisA image-level | AUROC: 92.0-93.5 on MVTec-AD | No clear effect |
| use SGNet instead of AA-CLIP | BTAD image-level · BTAD (image-level) | AUROC: 94.8 -> 92.9 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- AA-CLIP — AA-CLIP: Enhancing Zero-Shot Anomaly Detection via Anomaly-Aware CLIP (CVPR 2025) · better on 4 of 6 shared results
- FE-CLIP — FE-CLIP: Frequency Enhanced CLIP Model for Zero-Shot Anomaly Detection and Segmentation (ICCV 2025) · better on 5 of 6 shared results
- VCP-CLIP — VCP-CLIP: A visual context prompting model for zero-shot anomaly segmentation (ECCV 2024) · better on 1 of 1 shared result
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
- Proximity-CLIP: Text-Guided Semantic Proximity Learning for Zero-Shot Anomaly Detection (ECCV 2026 · similarity 0.89)
- Robust Zero-shot Anomaly Detection under Limited Auxiliary Anomaly Priors (ECCV 2026 · similarity 0.88)
- Towards Zero-Shot Anomaly Detection and Reasoning with Multimodal Large Language Models (CVPR 2025 · similarity 0.87)
- VETime: Vision Enhanced Zero-Shot Time Series Anomaly Detection (NeurIPS 2026 · similarity 0.85)
- A Unified Reasoning Framework for Holistic Zero-Shot Video Anomaly Analysis (NeurIPS 2025 · similarity 0.85)
- Towards Visual Discrimination and Reasoning of Real-World Physical Dynamics: Physics-Grounded Anomaly Detection (CVPR 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.
