TasksSotADatasetsPapersMethodsSubmitAbout
Papers With Code 2

A community resource for machine learning research: papers, code, benchmarks, and state-of-the-art results.

Explore

Notable BenchmarksAll SotADatasetsPapersMethods

Community

Submit ResultsAbout

Data sourced from the PWC Archive (CC-BY-SA 4.0). Built by the community, for the community.

Papers/Dinomaly: The Less Is More Philosophy in Multi-Class Unsup...

Dinomaly: The Less Is More Philosophy in Multi-Class Unsupervised Anomaly Detection

Jia Guo, Shuai Lu, Weihang Zhang, Fang Chen, Hongen Liao, Huiqi Li

2024-05-23CVPR 2025 1Unsupervised Anomaly DetectionPhilosophyAnomaly DetectionMulti-class Anomaly Detection
PaperPDFCode(official)Code

Abstract

Recent studies highlighted a practical setting of unsupervised anomaly detection (UAD) that builds a unified model for multi-class images. Despite various advancements addressing this challenging task, the detection performance under the multi-class setting still lags far behind state-of-the-art class-separated models. Our research aims to bridge this substantial performance gap. In this paper, we introduce a minimalistic reconstruction-based anomaly detection framework, namely Dinomaly, which leverages pure Transformer architectures without relying on complex designs, additional modules, or specialized tricks. Given this powerful framework consisted of only Attentions and MLPs, we found four simple components that are essential to multi-class anomaly detection: (1) Foundation Transformers that extracts universal and discriminative features, (2) Noisy Bottleneck where pre-existing Dropouts do all the noise injection tricks, (3) Linear Attention that naturally cannot focus, and (4) Loose Reconstruction that does not force layer-to-layer and point-by-point reconstruction. Extensive experiments are conducted across popular anomaly detection benchmarks including MVTec-AD, VisA, and Real-IAD. Our proposed Dinomaly achieves impressive image-level AUROC of 99.6%, 98.7%, and 89.3% on the three datasets respectively, which is not only superior to state-of-the-art multi-class UAD methods, but also achieves the most advanced class-separated UAD records.

Results

TaskDatasetMetricValueModel
Anomaly DetectionMPDDDetection AUROC97.2Dinomaly
Anomaly DetectionMPDDSegmentation AUROC99.1Dinomaly
Anomaly DetectionMVTec ADDetection AUROC99.77Dinomaly ViT-L (model-unified multi-class)
Anomaly DetectionMVTec ADSegmentation AP70.53Dinomaly ViT-L (model-unified multi-class)
Anomaly DetectionMVTec ADSegmentation AUPRO95.09Dinomaly ViT-L (model-unified multi-class)
Anomaly DetectionMVTec ADSegmentation AUROC98.54Dinomaly ViT-L (model-unified multi-class)
Anomaly DetectionMVTec ADDetection AUROC99.6Dinomaly ViT-B (model-unified multi-class)
Anomaly DetectionMVTec ADSegmentation AP69.29Dinomaly ViT-B (model-unified multi-class)
Anomaly DetectionMVTec ADSegmentation AUPRO94.79Dinomaly ViT-B (model-unified multi-class)
Anomaly DetectionMVTec ADSegmentation AUROC98.35Dinomaly ViT-B (model-unified multi-class)
Anomaly DetectionVisADetection AUROC98.9Dinomaly ViT-L (model-unified multi-class)
Anomaly DetectionVisAF1-Score96.1Dinomaly ViT-L (model-unified multi-class)
Anomaly DetectionVisASegmentation AUPRO94.8Dinomaly ViT-L (model-unified multi-class)
Anomaly DetectionVisASegmentation AUPRO (until 30% FPR)94.8Dinomaly ViT-L (model-unified multi-class)
Anomaly DetectionVisASegmentation AUROC99.1Dinomaly ViT-L (model-unified multi-class)
Anomaly DetectionMVTec ADDetection AUROC99.8Dinomaly-Large
Anomaly DetectionMVTec ADSegmentation AUROC98.5Dinomaly-Large
Anomaly DetectionMVTec ADDetection AUROC99.6Dinomaly-Base
Anomaly DetectionMVTec ADSegmentation AUROC98.4Dinomaly-Base

Related Papers

Multi-Stage Prompt Inference Attacks on Enterprise LLM Systems2025-07-213DKeyAD: High-Resolution 3D Point Cloud Anomaly Detection via Keypoint-Guided Point Clustering2025-07-17A Semi-Supervised Learning Method for the Identification of Bad Exposures in Large Imaging Surveys2025-07-17A Privacy-Preserving Framework for Advertising Personalization Incorporating Federated Learning and Differential Privacy2025-07-16Streaming 4D Visual Geometry Transformer2025-07-15Bridge Feature Matching and Cross-Modal Alignment with Mutual-filtering for Zero-shot Anomaly Detection2025-07-15Adversarial Activation Patching: A Framework for Detecting and Mitigating Emergent Deception in Safety-Aligned Transformers2025-07-12Towards High-Resolution 3D Anomaly Detection: A Scalable Dataset and Real-Time Framework for Subtle Industrial Defects2025-07-10