The reviewed record of science sign in
Pith

arxiv: 2111.07677 · v2 · pith:XMDP36FS · submitted 2021-11-15 · cs.CV

FastFlow: Unsupervised Anomaly Detection and Localization via 2D Normalizing Flows

Jiawei Yu , Ye Zheng , Xiang Wang , Wei Li , Yushuang Wu , Rui Zhao , Liwei Wu This is my paper

Reviewed by Pith T0 review T1 audit T2 compute T3 formal T4 kernel pith:XMDP36FSrecord.jsonopen to challenge →

classification cs.CV
keywords anomalydistributionfastflowdetectionfeaturefeaturesimageinference
0
0 comments X
read the original abstract

Unsupervised anomaly detection and localization is crucial to the practical application when collecting and labeling sufficient anomaly data is infeasible. Most existing representation-based approaches extract normal image features with a deep convolutional neural network and characterize the corresponding distribution through non-parametric distribution estimation methods. The anomaly score is calculated by measuring the distance between the feature of the test image and the estimated distribution. However, current methods can not effectively map image features to a tractable base distribution and ignore the relationship between local and global features which are important to identify anomalies. To this end, we propose FastFlow implemented with 2D normalizing flows and use it as the probability distribution estimator. Our FastFlow can be used as a plug-in module with arbitrary deep feature extractors such as ResNet and vision transformer for unsupervised anomaly detection and localization. In training phase, FastFlow learns to transform the input visual feature into a tractable distribution and obtains the likelihood to recognize anomalies in inference phase. Extensive experimental results on the MVTec AD dataset show that FastFlow surpasses previous state-of-the-art methods in terms of accuracy and inference efficiency with various backbone networks. Our approach achieves 99.4% AUC in anomaly detection with high inference efficiency.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 16 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. MATCH: Flow Matching for Multi-View Anomaly Detection

    cs.CV 2026-06 unverdicted novelty 7.0

    MATCH is the first flow matching method for multi-view anomaly detection, reporting SOTA results on Real-IAD and the first comprehensive evaluation on MANTA-Tiny while enabling real-time use by omitting the divergence term.

  2. Failure Identification in Imitation Learning Via Statistical and Semantic Filtering

    cs.RO 2026-04 unverdicted novelty 7.0

    FIDeL detects failures in imitation learning by building compact nominal representations via optimal transport, applying conformal prediction thresholds, and using VLMs for semantic filtering, outperforming baselines ...

  3. Novel Anomaly Detection Scenarios and Evaluation Metrics to Address the Ambiguity in the Definition of Normal Samples

    cs.CV 2026-04 unverdicted novelty 7.0

    Introduces scenarios and metrics for ambiguous normal samples in anomaly detection plus RePaste method achieving SOTA on the new metric on MVTec AD while retaining high AUROC and PRO.

  4. SubspaceAD: Training-Free Few-Shot Anomaly Detection via Subspace Modeling

    cs.CV 2026-02 accept novelty 7.0

    A training-free method fits PCA to DINOv2 features from few normal images and detects anomalies via reconstruction residual, reaching SOTA one-shot AUROC of 97.1% image-level on MVTec-AD and 93.2% on VisA.

  5. LogiCo: A Unified Framework for Logical and Structural Anomaly Detection

    cs.CV 2026-06 unverdicted novelty 6.0

    LogiCo is a unified framework using component-level feature reconstruction to detect both logical and structural anomalies, achieving SOTA results on four benchmarks with code publicly available.

  6. DeCoFlow: Structural Decomposition of Normalizing Flows for Continual Anomaly Detection

    cs.CV 2026-06 unverdicted novelty 6.0

    DeCoFlow decomposes normalizing flow subnets into frozen bases and low-rank adapters with alignment, auxiliary layers, and tail-aware loss to achieve continual anomaly detection with zero forgetting and few added parameters.

  7. Hypergraph Normal World Models for Logical Visual Anomaly Detection

    cs.CV 2026-06 unverdicted novelty 6.0

    Hypergraph model on DINOv2 tokens raises logical anomaly AUROC to 0.9279 on MVTec LOCO breakfast-box data by scoring an information quotient across local, relational, and hyperedge evidence.

  8. From Local Geometry to Global Pseudo Labeling for Robust Positive Unlabeled Learning under Covariate Shift

    cs.CV 2026-05 unverdicted novelty 6.0

    SPUNA leverages spectral neighborhood annotation on visual feature manifolds to enable robust PU learning for covariate shift detection, matching fully supervised performance.

  9. Subspace-Guided Feature Reconstruction for Unsupervised Anomaly Localization

    cs.CV 2023-09 unverdicted novelty 6.0

    A new framework learns low-dimensional subspaces from nominal samples and reconstructs target deep embeddings via self-expressive linear combinations to localize anomalies, claiming SOTA on three benchmarks.

  10. LiZAD: A Lightweight Zero-Shot Anomaly Detection Framework for Industrial Manufacturing

    cs.CV 2026-07 unverdicted novelty 5.0

    LiZAD reduces memory by 61.5%, parameters by 74.6%, and latency by 3.02x versus six prior ZSAD models while incurring only a 6.4% average P-AUROC drop on VisA, BTAD, MPDD, and MVTec-AD, with successful Jetson deployment.

  11. $\mu$Flow: Leveraging Average Images for Improving Generalisation of Deepfake Faces Detectors

    cs.CV 2026-06 unverdicted novelty 5.0

    μFlow trains a normalizing flow on averaged real-image features to detect deepfakes via likelihood in a fully out-of-distribution setting.

  12. IndusAgent: Reinforcing Open-Vocabulary Industrial Anomaly Detection with Agentic Tools

    cs.CV 2026-05 unverdicted novelty 5.0

    IndusAgent achieves state-of-the-art zero-shot performance on industrial anomaly benchmarks by using a custom Indus-CoT dataset, dynamic tool orchestration, and gated RL to optimize anomaly classification, localizatio...

  13. ZSG-IAD: A Multimodal Framework for Zero-Shot Grounded Industrial Anomaly Detection

    cs.CV 2026-04 unverdicted novelty 5.0

    ZSG-IAD is a zero-shot multimodal system that uses language-guided two-hop grounding and rule-based reinforcement learning to produce anomaly masks and explainable reports from industrial sensor data.

  14. MambaADv2: Evolving Duality-enhanced State Space Model for Unsupervised Anomaly Detection

    cs.CV 2026-06 unverdicted novelty 4.0

    MambaADv2 evolves Mamba state space models with hybrid blocks, frequency convolutions, and adaptive scanning for improved unsupervised anomaly detection.

  15. Leveraging Unsupervised Learning for Cost-Effective Visual Anomaly Detection

    cs.CV 2024-09 unverdicted novelty 3.0

    Unsupervised anomaly detection with pre-trained Anomalib models achieves F1 macro score over 0.95 on Raspberry Pi using 10 images and 90 seconds training time.

  16. Real-Time Industrial Defect Detection on Edge Hardware Using Fine-Tuned YOLOv8: A Systematic Benchmark on the NEU Surface Defect Database and MVTec AD with Automotive & Battery Manufacturing Extensions

    cs.CV 2026-06 unverdicted novelty 2.0

    Fine-tuned YOLOv8 with TensorRT/OpenVINO optimizations reaches over 120 FPS and 98.5% mAP on NEU, MVTec AD, and custom automotive defect datasets for edge deployment.