Flow Mismatching: Unsupervised Anomaly Detection via Velocity Discrepancies in Flow Matching Models
Quick Answer
Flow Mismatching introduces an unsupervised anomaly detection method that leverages geometric dynamics to identify discrepancies in velocity between normal and anomalous flows.
Quick Take
Flow Mismatching introduces an unsupervised anomaly detection method that leverages geometric dynamics to identify discrepancies in velocity between normal and anomalous flows. By analyzing pixel-wise heatmaps and scores from trained flow matching models on datasets like MVTec-AD and VisA, it outperforms existing reconstruction-based methods without requiring test-time optimization.
Key Points
- Proposed method avoids reconstruction-based paradigms for anomaly detection.
- Utilizes geometric dynamics to assess velocity discrepancies in flow matching.
- Achieves superior performance on MVTec-AD and VisA datasets.
- Generates pixel-wise heatmaps without requiring additional calibration.
- Identifies anomalies through local velocity disagreements along affine paths.
Paper Resources
Article Content
From source RSS / original summaryarXiv:2605. 23070v1 Announce Type: new Abstract: We propose Flow Mismatching, an unsupervised anomaly detection method that deliberately avoids reconstruction-based paradigms. Instead, we treat flow matching as geometric dynamics and leverage a key insight: anomalies occur at places where the learned normal flow disagrees with the geometric path toward a test image.
Given a flow matching model trained only on normal images, we probe its learned velocity field along affine paths from Gaussian noise to a target image. Along each path, we compare the model-predicted velocity, which follows normal generative dynamics, with the geometric velocity toward the target, which includes any anomalous content. Anomalies induce strong local disagreement between these velocities.
Aggregating the mismatch over different time steps and multiple paths yields pixel-wise heatmaps and image-level scores without test-time optimization, feature memories, or additional calibration. Our analysis shows that the population mismatch decomposes into an irreducible denoising term and a Fisher-divergence term between the test-path and normal-path score functions, which identifies the score-gap component that drives anomaly separation and explains the effectiveness of robust path aggregation.
Extensive experiments on MVTec-AD and VisA demonstrate superior performance compared with SOTA reconstruction-based and recent flow matching-based approaches.
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.CV
See more →ProMoE-FL: Prototype-conditioned Mixture of Experts for Multimodal Federated Learning with Missing Modalities
ProMoE-FL introduces a Prototype-conditioned Mixture-of-Experts framework for multimodal federated learning, effectively addressing missing modalities. It outperforms existing methods on four chest X-ray datasets, demonstrating superior feature synthesis capabilities in both homogeneous and heterogeneous settings.