CFLOW-AD: Real-Time Unsupervised Anomaly Detection with Localization via Conditional Normalizing Flows

By Denis A. Gudovskiy, Shun Ishizaka, Kazuki Kozuka. IEEE Workshop/Winter Conference on Applications of Computer Vision, 2021

656 citations recorded by the source corpus.

Unsupervised anomaly detection with localization has many practical applications when labeling is infeasible and, moreover, when anomaly examples are completely missing in the train data. While recently proposed models for such data setup achieve high accuracy metrics, their complexity is a limiting factor for real-time processing. In this paper, we propose a real-time model and analytically derive its relationship to prior methods. Our CFLOW-AD model is based on a conditional normalizing flow frame- work adopted for anomaly detection with localization. In particular, CFLOW-AD consists of a discriminatively pretrained encoder followed by a multi-scale generative de- coders where the latter explicitly estimate likelihood of the encoded features. Our approach results in a computationally and memory-efficient model: CFLOW-AD is faster and smaller by a factor of 10× than prior state-of-the-art with the same input setting. Our experiments on the MVTec dataset show that CFLOW-AD outperforms previous methods by 0.36% AUROC in detection task, by 1.12% AUROC and 2.5% AUPRO in localization task, respectively. We open-source our code with fully reproducible experiments1.

Subfields: Artificial Intelligence, Computer Networks and Communications, Signal Processing

View Paper: https://www.semanticscholar.org/paper/fc086bf5f6d1627153b68abdd5a4450e141b4ca3

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Rui Zhao
2. Ye Zheng
3. Kilian Batzner
4. Jiulong Shan
5. Zhikang Liu
6. Ming Tang
7. Xi Li
8. Lars Heckler

In Computer Networks and Communications

1. Rui Zhao
2. Ye Zheng
3. Kilian Batzner
4. Zhikang Liu
5. Xi Li
6. Lars Heckler
7. Jiawei Yu1
8. Xiang Wang

In Signal Processing

1. Rui Zhao
2. Kilian Batzner
3. Lars Heckler
4. Rebecca König

Citations in the corpus

• AnomalyGPT: Detecting Industrial Anomalies using Large Vision-Language Models (2023, background): “Cflow-AD [9] and PyramidFlow [13] use normalizing flows to project normal samples onto a Gaussian distribution.”
• FastFlow: Unsupervised Anomaly Detection and Localization via 2D Normalizing Flows (2021, methodology): “Recently, some works (Rudolph, Wandt, and Rosenhahn 2021; Gudovskiy, Ishizaka, and Kozuka 2021) began to use normalizing flow (Kingma and Dhariwal 2018) to estimate distribution.”
• Deep Industrial Image Anomaly Detection: A Survey (2023, neutral): “In addition, CFlow-AD [53] analyzes in depth why the multivariate Gaussian assumption is a reasonable prior in earlier models and why the more general NF framework aims to converge to similar results with less computation.”
• EfficientAD: Accurate Visual Anomaly Detection at Millisecond-Level Latencies (2023, methodology): “Common methods include multivariate Gaussian distributions [16, 30, 48], Gaussian Mixture Models [38, 72], Normalizing Flows [22,47,51,52,68], and the k-Nearest Neighbor (kNN) algorithm [14, 39, 40, 50].”
• SimpleNet: A Simple Network for Image Anomaly Detection and Localization (2023, methodology): “CS-Flow [24], CFLOW-AD [12], and DifferNet [23] propose to transform the normal feature distribution into Gaussian distribution via normalizing flow (NF) [20].”
• DeSTSeg: Segmentation Guided Denoising Student-Teacher for Anomaly Detection (2022, background): “In addition, the downsampling implementations are not standardized either [13,25,37], resulting in the varying ground truth and unfair evaluation.”