Natural Synthetic Anomalies for Self-supervised Anomaly Detection and Localization

Schlüter, Hannah M.; Tan, Jeremy; Hou, Benjamin; Kainz, Bernhard

doi:10.1007/978-3-031-19821-2_27

Cited by 62 publications

(55 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…If pairwise applied, the allocation of image pairs is randomly drawn from the image batch. Second, we take patches from different locations of the source image and interpolate into also different location inside the target image, hence, we latch on the patch drawing by NSA [22]. Third, we overcome the current limitation of PII and PII-based anomaly generation methods regarding the grade of abnormality of the interpolated patches.…”

Section: Methodsmentioning

confidence: 99%

“…Poisson Image Interpolation (PII) [26] overcomes sharp discontinuities with Poisson editing as interpolation strategy and generates more organic and subtle outliers. Natural Synthetic Anomalies (NSA) [22] are introduced by rescaling, shifting and a new Gammadistribution-based patch shape sampling without the use of interpolation factors for an end-to-end model for anomaly detection.…”

Section: Related Workmentioning

confidence: 99%

“…However, such synthesis strategies feature obvious discontinuities. [26,22] solve the discontinuity problem by using Poisson image editing, but the resulting anomalies can be so subtle that they may represent variations of the normal class rather than true anomalies and these approaches do not provide prediction confidence estimates. Therefore we propose a new approach to model the ID space and make the following contributions:…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Confidence-Aware and Self-Supervised Image Anomaly Localisation

Müller¹,

Baugh²,

Tan³

et al. 2023

Preprint

View full text Add to dashboard Cite

Universal anomaly detection still remains a challenging problem in machine learning and medical image analysis. It is possible to learn an expected distribution from a single class of normative samples, e.g., through epistemic uncertainty estimates, auto-encoding models, or from synthetic anomalies in a self-supervised way. The performance of self-supervised anomaly detection approaches is still inferior compared to methods that use examples from known unknown classes to shape the decision boundary. However, outlier exposure methods often do not identify unknown unknowns. Here we discuss an improved self-supervised single-class training strategy that supports the approximation of probabilistic inference with loosen feature locality constraints. We show that up-scaling of gradients with histogram-equalised images is beneficial for recently proposed self-supervision tasks. Our method is integrated into several out-of-distribution (OOD) detection models and we show evidence that our method outperforms the state-of-the-art on various benchmark datasets. Source code will be publicly available by the time of the conference.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Confidence-Aware and Self-Supervised Image Anomaly Localisation

Müller¹,

Baugh²,

Tan³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…DSR [85] L2, Focal -The paper generates abnormal samples in feature level and perform better than DRAEM. NSA [86] L2, Cross-Entropy -The paper generates abnormal samples by pasting parts of other normal samples, which is the SOTA method without extra data. SSPCAB [87] L2 -The paper designs a "plug and play" self-supervised block to improve the reconstruction ability of many methods.…”

Section: Vggmentioning

confidence: 99%

“…By sampling the learned quantized feature space at the feature level, the near-in-distribution anomalies are generated in a controlled way. NSA [86] does not use external data for data data augmentation and adopts more data augmentation methods, allowing it to outperform all previous methods that learned without utilizing additional datasets. In contrast to other methods that attempt to reconstruct abnormal images into normal images, Bauer [82] proposes reconstructing the abnormal areas of the image so that they deviate from the original image's appearance.…”

Section: Autoencodermentioning

confidence: 99%

Deep Industrial Image Anomaly Detection: A Survey

Liu¹,

Xie²,

Wang³

et al. 2023

Preprint

View full text Add to dashboard Cite

The recent rapid development of deep learning has laid a milestone in industrial Image Anomaly Detection (IAD). In this paper, we provide a comprehensive review of deep learning-based image anomaly detection techniques, from the perspectives of neural network architectures, levels of supervision, loss functions, metrics and datasets. In addition, we extract the new setting from industrial manufacturing and review the current IAD approaches under our proposed our new setting. Moreover, we highlight several opening challenges for image anomaly detection. The merits and downsides of representative network architectures under varying supervision are discussed. Finally, we summarize the research findings and point out future research directions. More resources are available at https://github.com/M-3LAB/awesome-industrial-anomaly-detection.

show abstract