Patch SVDD: Patch-level SVDD for Anomaly Detection and Segmentation

By Jihun Yi, Sungroh Yoon. Asian Conference on Computer Vision, 2020

520 citations recorded by the source corpus.

In this paper, we address the problem of image anomaly detection and segmentation. Anomaly detection involves making a binary decision as to whether an input image contains an anomaly, and anomaly segmentation aims to locate the anomaly on the pixel level. Support vector data description (SVDD) is a long-standing algorithm used for an anomaly detection, and we extend its deep learning variant to the patch-based method using self-supervised learning. This extension enables anomaly segmentation and improves detection performance. As a result, anomaly detection and segmentation performances measured in AUROC on MVTec AD dataset increased by 9.8% and 7.0%, respectively, compared to the previous state-of-the-art methods. Our results indicate the efficacy of the proposed method and its potential for industrial application. Detailed analysis of the proposed method offers insights regarding its behavior, and the code is available online.

Subfields: Artificial Intelligence, Computer Networks and Communications, Endocrinology

View Paper: https://www.semanticscholar.org/paper/62b77e5cb85fc61b84edd532f6d65714be152596

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

In Computer Networks and Communications

In Endocrinology

Citations in the corpus

Student-Teacher Feature Pyramid Matching for Anomaly Detection (2021, methodology): “…two trends to achieve performance gains: (1) by pre-trained models, with a Wide-ResNet50 × 2 network [44], SPADE reports very competitive scores; (2) by self-training techniques, Cut-Paste [21] and Patch-SVDD [43] show this potential through designing proper pretext tasks for feature learning.”
A Unified Model for Multi-class Anomaly Detection (2022, background): “US [3] PSVDD [15] PaDiM [5] CutPaste [10] FCDD [11] MKD [12] DRAEM [16] Ours FLOPs(G) 60.”
FastFlow: Unsupervised Anomaly Detection and Localization via 2D Normalizing Flows (2021, methodology): “…networks to obtain the features of normal images and model the distribution with some statistical methods, then detect the abnormal samples that have different distributions (Bergman and Hoshen 2020; Rippel, Mertens, and Merhof 2021; Yi and Yoon 2020; Cohen and Hoshen 2020; Defard et al. 2020).”
Deep Industrial Image Anomaly Detection: A Survey (2023, unknown): “PatchSVDD [18] divides the image into uniform patches and sends them to the model for training, which significantly enhances the model’s ability to detect anomalies.”
CFLOW-AD: Real-Time Unsupervised Anomaly Detection with Localization via Conditional Normalizing Flows (2021, methodology): “Similarly to [37, 7, 42], this approach relies on ImageNetpretrained feature extractor with multi-scale pyramid pooling.”
Anomaly Detection via Reverse Distillation from One-Class Embedding (2022, methodology): “To cope with high-dimensional data, DeepSVDD [31] and PatchSVDD [43] estimate data representations through deep networks.”
SPot-the-Difference Self-Supervised Pre-training for Anomaly Detection and Segmentation (2022, methodology): “AU-ROC is the most widely used metric for anomaly detection tasks [14,33,44].”
DeSTSeg: Segmentation Guided Denoising Student-Teacher for Anomaly Detection (2022, background): “Besides, the memorybased approaches [7,20,25,35] build a memory bank of normal data in training.”
AnomalyGPT: Detecting Industrial Anomalies using Large Vision-Language Models (2023, background): “Approaches such as PatchSVDD [31] aim to find a hypersphere that tightly encapsulates normal samples.”
WinCLIP: Zero-/Few-Shot Anomaly Classification and Segmentation (2023, unknown): “For classification, we report (a) Area Under the Receiver Operating Characteristic (AUROC) following the literature [8,31,48], as well as (b) Area Under the Precision-Recall curve (AUPR) and (c) F1-score at optimal threshold (F1-max) for a clearer view against potential data imbalance [57]).”
CFA: Coupled-Hypersphere-Based Feature Adaptation for Target-Oriented Anomaly Localization (2022, background): “Previous studies [18] and [23] that learned the distribution of target datasets by introducing the hypersphere concept still had a problem in not clearly understanding normal features because they did not use a memory bank.”
CutPaste: Self-Supervised Learning for Anomaly Detection and Localization (2021, methodology): “With a finer-grained 3-way classification to leverage different scale of CutPaste, we achieve the best 95.2 AUC, which surpasses existing works on learning from scratch, such as P-SVDD [61] (92.1 AUC).”
Towards Total Recall in Industrial Anomaly Detection (2021, background): “As in [10,14,56], images are resized and center cropped to 256 × 256 and 224 × 224 , respectively.”
Self-Supervised Predictive Convolutional Attentive Block for Anomaly Detection (2021, background): “Anomaly detection is an important task with a broad set of applications ranging from industrial inspection (finding defects of objects or materials on industrial production lines) [5, 7, 10, 15, 36, 56, 62, 76] to public security (detecting abnormal events such as traffic accidents, fights, explosio”
DRÆM – A discriminatively trained reconstruction embedding for surface anomaly detection (2021, methodology): “Recently Patch-based one-class classiﬁcation methods have been considered for surface anomaly detection [30].”
PaDiM: a Patch Distribution Modeling Framework for Anomaly Detection and Localization (2020, unknown): “Type Reconstruction-based methods Embedding similarity based methods Our methods Model AE simm [1], [2], [9] AE L2 [1], [2] VAE Student [2] Patch SVDD [4] SPADE [5] PaDiMR18-Rd100 PaDiMWR50-Rd550 Carpet (87, 64.”