CutPaste: Self-Supervised Learning for Anomaly Detection and Localization

Li, Chun-Liang; Sohn, Kihyuk; Yoon, Jinsung; Pfister, Tomas

doi:10.1109/cvpr46437.2021.00954

Cited by 501 publications

(345 citation statements)

References 18 publications

Supporting

Mentioning

276

Contrasting

Order By: Relevance

“…Only the approach described in [14] analyzes the advantages of transferring features learned by image inpainting to other tasks, and the authors have achieved competitive results with other models that use supervised pretraining. Authors of [28] show that selfsupervised learning can be used for anomaly detection by learning deep representations and then building a generative one-class classifier on learned representations. Another approach described in [29] proposes a deraining method using an unsupervised deraining generative adversarial network that resolves issues of previous approaches by introducing self-supervised constraints from the intrinsic statistics of unpaired rainy and clean images.…”

Section: Related Workmentioning

confidence: 99%

GAN-Based Image Colorization for Self-Supervised Visual Feature Learning

Treneska

Zdravevski

Pires

et al. 2022

Sensors

View full text Add to dashboard Cite

Large-scale labeled datasets are generally necessary for successfully training a deep neural network in the computer vision domain. In order to avoid the costly and tedious work of manually annotating image datasets, self-supervised learning methods have been proposed to learn general visual features automatically. In this paper, we first focus on image colorization with generative adversarial networks (GANs) because of their ability to generate the most realistic colorization results. Then, via transfer learning, we use this as a proxy task for visual understanding. Particularly, we propose to use conditional GANs (cGANs) for image colorization and transfer the gained knowledge to two other downstream tasks, namely, multilabel image classification and semantic segmentation. This is the first time that GANs have been used for self-supervised feature learning through image colorization. Through extensive experiments with the COCO and Pascal datasets, we show an increase of 5% for the classification task and 2.5% for the segmentation task. This demonstrates that image colorization with conditional GANs can boost other downstream tasks’ performance without the need for manual annotation.

show abstract

Section: Related Workmentioning

confidence: 99%

GAN-Based Image Colorization for Self-Supervised Visual Feature Learning

Treneska

Zdravevski

Pires

et al. 2022

Sensors

View full text Add to dashboard Cite

show abstract

“…We experiment with three image datasets: CIFAR-10, Fashion-MNIST, and MVTEC (Bergmann et al, 2019). These have been used in virtually all deep anomaly detection papers published at top-tier venues (Ruff et al, 2018;Golan & El-Yaniv, 2018;Hendrycks et al, 2019;Bergman & Hoshen, 2020;Li et al, 2021), and we adopt these papers' experimental protocol here, as detailed below.…”

Section: Experiments On Image Datamentioning

confidence: 99%

Latent Outlier Exposure for Anomaly Detection with Contaminated Data

Qiu¹,

Li²,

Kloft³

et al. 2022

Preprint

View full text Add to dashboard Cite

Anomaly detection aims at identifying data points that show systematic deviations from the majority of data in an unlabeled dataset. A common assumption is that clean training data (free of anomalies) is available, which is often violated in practice. We propose a strategy for training an anomaly detector in the presence of unlabeled anomalies that is compatible with a broad class of models. The idea is to jointly infer binary labels to each datum (normal vs. anomalous) while updating the model parameters. Inspired by outlier exposure (Hendrycks et al., 2018) that considers synthetically created, labeled anomalies, we thereby use a combination of two losses that share parameters: one for the normal and one for the anomalous data. We then iteratively proceed with block coordinate updates on the parameters and the most likely (latent) labels. Our experiments with several backbone models on three image datasets, 30 tabular data sets, and a video anomaly detection benchmark showed consistent and significant improvements over the baselines.

show abstract

“…25 Crafting synthetic outliers can be considered as self-supervised learning, since no manual labelling is required. In CutPaste, 7 the authors propose to transform any normal training image into an anomalous one by locally incorporating a transformed patch coming from any of the other training images. This allows to learn finer features, which will be more specific to the task-at-hand than outliers coming from a different dataset.…”

Section: "Unsupervised" Learning For Anomaly Detectionmentioning

confidence: 99%

“…We first assess in section 4.3 the impact of defective training parts on the performances of these detectors, and show that the latter are already quite robust to pollution compared to other methods such as CutPaste. 7 2. We then propose in section 4.4 a simple yet efficient refinement strategy to remove polluted samples from the training data.…”

Section: Introductionmentioning

confidence: 99%

Data refinement for fully unsupervised visual inspection using pre-trained networks

Cordier¹,

Missaoui²,

Gutierrez³

2022

Preprint

View full text Add to dashboard Cite

Anomaly detection has recently seen great progress in the field of visual inspection. More specifically, the use of classical outlier detection techniques on features extracted by deep pre-trained neural networks have been shown to deliver remarkable performances on the MVTec Anomaly Detection (MVTec AD) dataset. However, like most other anomaly detection strategies, these pre-trained methods assume all training data to be normal. As a consequence, they cannot be considered as fully unsupervised. There exists to our knowledge no work studying these pre-trained methods under fully unsupervised setting. In this work, we first assess the robustness of these pre-trained methods to fully unsupervised context, using polluted training sets (i.e. containing defective samples), and show that these methods are more robust to pollution compared to methods such as CutPaste. We then propose SROC, a Simple Refinement strategy for One Class classification. SROC enables to remove most of the polluted images from the training set, and to recover some of the lost AUC. We further show that our simple heuristic competes with, and even outperforms much more complex strategies from the existing literature.

show abstract

CutPaste: Self-Supervised Learning for Anomaly Detection and Localization

Cited by 501 publications

References 18 publications

GAN-Based Image Colorization for Self-Supervised Visual Feature Learning

GAN-Based Image Colorization for Self-Supervised Visual Feature Learning

Latent Outlier Exposure for Anomaly Detection with Contaminated Data

Data refinement for fully unsupervised visual inspection using pre-trained networks

Contact Info

Product

Resources

About