Semantic Iterative Reconstruction: One-Shot Universal Anomaly Detection

2026-03-24

AI Summary

This work addresses the challenges of unsupervised medical anomaly detection, which is often hindered by the scarcity of normal samples and limited cross-modal generalization. The authors propose a semantic iterative reconstruction framework that establishes a novel one-shot universal anomaly detection paradigm: using only a single normal sample per heterogeneous dataset, the method trains a unified model capable of generalizing across diverse multimodal medical imaging data. Leveraging a pretrained teacher encoder to extract multiscale deep features, the approach integrates a compact up/down-sampling decoder with multi-round iterative refinement to strengthen normality priors in feature space. Evaluated on nine medical benchmarks, the method achieves state-of-the-art performance across four evaluation settings—including one-shot and full-sample scenarios—significantly outperforming existing approaches.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Multi-modal Vision
• Data Mining & Knowledge Management: Anomaly/Outlier Detection

Application Category

• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Unsupervised medical anomaly detection is severely limited by the scarcity of normal training samples. Existing methods typically train dedicated models for each dataset or disease, requiring hundreds of normal images per task and lacking cross-modality generalization. We propose Semantic Iterative Reconstruction (SIR), a framework that enables a single universal model to detect anomalies across diverse medical domains using extremely few normal samples. SIR leverages a pretrained teacher encoder to extract multi-scale deep features and employs a compact up-then-down decoder with multi-loop iterative refinement to enforce robust normality priors in deep feature space. The framework adopts a one-shot universal design: a single model is trained by mixing exactly one normal sample from each of nine heterogeneous datasets, enabling effective anomaly detection on all corresponding test sets without task-specific retraining. Extensive experiments on nine medical benchmarks demonstrate that SIR achieves state-of-the-art under all four settings -- one-shot universal, full-shot universal, one-shot specialized, and full-shot specialized -- consistently outperforming previous methods. SIR offers an efficient and scalable solution for multi-domain clinical anomaly detection. Code is available at https://github.com/jusufzn212427/sir4ad.

Problem

• unsupervised medical anomaly detection
• data scarcity
• cross-modality generalization
• one-shot learning
• universal anomaly detection

Innovation

• one-shot learning
• universal anomaly detection
• semantic iterative reconstruction
• medical image analysis
• unsupervised anomaly detection