Attribute Restoration Framework for Anomaly Detection

Fei, Ye; Huang, Chaoqin; Cao, Jinkun; Li, Maosen; Zhang, Ya; Lu, Cewu

doi:10.1109/tmm.2020.3046884

Cited by 108 publications

(65 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Comparison to state of the art. We compared our method with ITAE [23], AESC [24] and SPADE [25]. The proposed method achieves state-of-the-art results as shown as it can be seen in Table 1, it outperforms the state of the art by about 10% on average.…”

Section: Methodsmentioning

confidence: 82%

See 1 more Smart Citation

Unsupervised Variability Normalization For Anomaly Detection

Artola

Kolodziej²,

Morel

et al. 2021

2021 IEEE International Conference on Image Processing (ICIP)

View full text Add to dashboard Cite

Anomaly detectors are necessary to automatize industrial quality control. However, crafting such detectors is difficult due to the complexity and variability of the object even when working only with rigid objects. We show that adding a deep learning normalization step as a preprocessing step to model based detectors allows for better and more robust detections. This self-supervised normalization neural network is trained on non-anomalous data only. The proposed preprocessing method, followed by an automatic detector, achieves state-ofthe-art results on rigid objects from the MvTec dataset.

show abstract

Section: Methodsmentioning

confidence: 82%

“…More recently, sophisticated deep learning based methods have also been used to try to solve this problem. ITEA [23] proceeds on grey level images, normalizes their orientation and trains an auto-encoder on anomaly free images. AESC [24] is similar except that here the network is trained to denoise the image from a stain noise.…”

Section: Related Workmentioning

confidence: 99%

Unsupervised Variability Normalization For Anomaly Detection

Artola

Kolodziej²,

Morel

et al. 2021

2021 IEEE International Conference on Image Processing (ICIP)

View full text Add to dashboard Cite

show abstract

“…Mayr 等人 [150] 基于 Resnet50 的骨干网络设计二分类网络, 基于 CAM 的思想, 利用分类结果生成缺陷的激活图, 从而得到分割结果. Venkataramanan 等人 [120] 述. Fei 等人 [123] 采用属性擦除的方式, 让模型可以恢复正常图像的颜色、方向等属性, 并将异常像素复原. Salehi 等人 [114] 利用拼图还原的代理任务, 使网络具有恢复正常图像的能力.…”

Section: 弱监督学习unclassified

Survey on industrial defect detection with deep learning

Luo¹,

Cai²,

Yang³

et al. 2022

Sci. Sin.-Inf.

View full text Add to dashboard Cite

Deep-learning based industrial defect detection methods play an increasingly important role in intelligent manufacturing, as they provide compelling benefits in reducing the cost spent on manual product inspection and meanwhile, improving inspection accuracy and efficiency. They have been widely used in various manufacturing and O&M applications such as automated inspection, smart patrol and quality controlling. This survey aims to make a comprehensive introduction of industrial defect detection, which mainly spans its definition, difficulties, challenges, mainstream methods, open datasets and evaluation protocols, so as to help researchers gather a quick and broad understanding. Specifically, we firstly introduce some background knowledge. Secondly, based on the difference of the provided annotations of different datasets in practical scenarios, we categorize most methods into three task settings: known defects, unknown defects, and few-shot defects. We give more details over these methods and illustrate a thorough analysis. We expound the connections between different algorithms and actual demands to provide a clear image of how different algorithms evolve. Thirdly, this paper summarizes some useful strategies that can effectively improve defect detection performance. Finally, based on our understanding of this area, we conclude several limitations of existing methods in practical applications as well as several directions of future research that embrace further efforts.

show abstract

“…A stable model can ensure the performance of the final model is not unacceptable. The stability of model performance is mainly reflected in three aspects 32 To assess the stability of our method, we monitor the UAD performance by AUROC of our method and the CAE during the training phase. Figure 3 shows the AUROC in different training epochs.…”

Section: Model Stabilitymentioning

confidence: 99%

Improved autoencoder for unsupervised anomaly detection

Cheng

Wang

Zhang

et al. 2021

Int J of Intelligent Sys

View full text Add to dashboard Cite

Deep autoencoder-based methods are the majority of deep anomaly detection. An autoencoder learning on training data is assumed to produce higher reconstruction error for the anomalous samples than the normal samples and thus can distinguish anomalies from normal data. However, this assumption does not always hold in practice, especially in unsupervised anomaly detection, where the training data is anomaly

show abstract

Attribute Restoration Framework for Anomaly Detection

Cited by 108 publications

References 42 publications

Unsupervised Variability Normalization For Anomaly Detection

Unsupervised Variability Normalization For Anomaly Detection

Survey on industrial defect detection with deep learning

Improved autoencoder for unsupervised anomaly detection

Contact Info

Product

Resources

About