Research map / Vision-language and multimodal / CLIP and zero-shot
DictAS: A Framework for Class-Generalizable Few-Shot Anomaly Segmentation via Dictionary Lookup
Zhen Qu, Xian Tao, Xinyi Gong, ShiChen Qu, Xiaopei Zhang, Xingang Wang, Fei Shen, Zhengtao Zhang, Mukesh Prasad, Guiguang Ding
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 Query Generator (ablation) | DictAS · MVTecAD 4-shot | pixel AP: 63.5 -> 66.8 | Improved |
| add Key Generator (ablation) | DictAS · MVTecAD 4-shot | pixel AP: 63.8 -> 66.8 | Improved |
| add Value Generator (ablation) | DictAS · MVTecAD 4-shot | pixel AP: 64.2 -> 66.8 | Improved |
| add Contrastive Query Constraint (ablation) | DictAS · MVTecAD 4-shot | pixel AP: 64.6 -> 66.8 | Improved |
| add Text Alignment Constraint (ablation) | DictAS · MVTecAD 4-shot | pixel AP: 65.0 -> 66.8 | Improved |
| add both query discrimination regularizers (ablation) | DictAS · MVTecAD 4-shot | pixel AP: 63.7 -> 66.8 | Improved |
| use sparse lookup instead of maximum lookup (ablation) | DictAS · MVTecAD 16-shot | pixel AP: 60.6 -> 68.5 | Improved |
| use sparse lookup instead of dense lookup (ablation) | DictAS · MVTecAD 16-shot | pixel AP: 63.8 -> 68.5 | Improved |
| use sparse lookup instead of dense lookup (ablation) | DictAS · MVTecAD 1-shot | pixel AP: 60.2 -> 61.1 | Improved |
| raise input resolution from 336x336 to 518x518 (ablation) | DictAS · MVTecAD 4-shot | pixel AP: 66.8 -> 68.7 | Improved |
| use ViT-L-14-336 instead of ViT-B-16 backbone (ablation) | DictAS · MVTecAD 4-shot | pixel AP: 64.8 -> 66.8 | Improved |
| train on natural-scene Ade20K instead of VisA (ablation) | DictAS · MVTecAD 4-shot | pixel AP: 66.8 -> 66.5 | No clear effect |
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.
- WinCLIP — WinCLIP: Zero-/Few-Shot Anomaly Classification and Segmentation (CVPR 2023) · 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
- MetaUAS: Universal Anomaly Segmentation with One-Prompt Meta-Learning (NeurIPS 2024 · similarity 0.88)
- AnomalyCLIP: Object-agnostic Prompt Learning for Zero-shot Anomaly Detection (ICLR 2024 · similarity 0.87)
- UniVAD: A Training-free Unified Model for Few-shot Visual Anomaly Detection (CVPR 2025 · similarity 0.87)
- Adapting Visual-Language Models for Generalizable Anomaly Detection in Medical Images (CVPR 2024 · similarity 0.87)
- Towards Zero-Shot Anomaly Detection and Reasoning with Multimodal Large Language Models (CVPR 2025 · similarity 0.87)
- Toward Generalist Anomaly Detection via In-context Residual Learning with Few-shot Sample Prompts (CVPR 2024 · 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.
