UniFormaly: Towards Task-Agnostic Unified Framework for Visual Anomaly Detection

2023/07/24 by Yujin Lee, Lee, Yujin, Harin Lim +3 · 5 citations
Computer Science · Medicine · #Anomaly Detection Techniques and Applications #Artificial Intelligence (cs.AI) #COVID-19 diagnosis using AI #Computer Vision and Pattern Recognition (cs.CV) #Data-Driven Disease Surveillance #FOS: Computer and information sciences

paper · pdf · doi:10.48550/arxiv.2307.12540

openalex publication_date 2023/07/24 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Visual anomaly detection aims to learn normality from normal images, but existing approaches are fragmented across various tasks: defect detection, semantic anomaly detection, multi-class anomaly detection, and anomaly clustering. This one-task-one-model approach is resource-intensive and incurs high maintenance costs as the number of tasks increases. We present UniFormaly, a universal and powerful anomaly detection framework. We emphasize the necessity of our off-the-shelf approach by pointing out a suboptimal issue in online encoder-based methods. We introduce Back Patch Masking (BPM) and top k-ratio feature matching to achieve unified anomaly detection. BPM eliminates irrelevant background regions using a self-attention map from self-supervised ViTs. This operates in a task-agnostic manner and alleviates memory storage consumption, scaling to tasks with large-scale datasets. Top k-ratio feature matching unifies anomaly levels and tasks by casting anomaly scoring into multiple instance learning. Finally, UniFormaly achieves outstanding results on various tasks and datasets. Codes are available at https://github.com/YoojLee/Uniformaly.

Cited by

• ProtoAnomalyNCD: Prototype Learning for Multi-class Novel Anomaly Discovery in Industrial Scenarios
• A Lightweight 3D Anomaly Detection Method with Rotationally Invariant Features
• Image-Intrinsic Priors for Integrated Circuit Defect Detection and Novel Class Discovery via Self-Supervised Learning
• Generalist Multi-Class Anomaly Detection via Distillation to Two Heterogeneous Student Networks
• Multi-class Image Anomaly Detection for Practical Applications: Requirements and Robust Solutions

Related

• AnomalyMoE: Towards a Language-free Generalist Model for Unified Visual Anomaly Detection
• UniADC: A Unified Framework for Anomaly Detection and Classification
• GeneralAD: Anomaly Detection Across Domains by Attending to Distorted Features
• Anomaly Detection by Adapting a pre-trained Vision Language Model
• From Benchmarks to Reality: Advancing Visual Anomaly Detection by the VAND 3.0 Challenge