Anomaly Detection via Reverse Distillation from One-Class Embedding

Deng, Hanqiu; Li, Xingyu

doi:10.1109/cvpr52688.2022.00951

Cited by 293 publications

(165 citation statements)

References 29 publications

Supporting

Mentioning

163

Contrasting

Order By: Relevance

“…The paper adds another teacher-student pair to get different feature reconstruction results. RD4AD [14] Cosine Similarity ResNet The paper designs the teacher-student model of reverse distillation in a similar way to reconstruction. IKD [15] Context Similarity ResNet The paper adds context similarity loss and adaptive hard sample mining module to prevent overfitting.…”

Section: Resnetmentioning

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

Section: Resnetmentioning

confidence: 99%

Deep Industrial Image Anomaly Detection: A Survey

Liu¹,

Xie²,

Wang³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…In terms of reconstruction-based methods, models are trained to learn features that reconstruct the original data, which identifies poorly reconstructed samples as anomalies. Accordingly, extensions based on AEs [5,7,10,14,20,31] or GANs [2,22,23,30,32,36] have been proposed. In terms of embedding similarity-based methods, the latent features of nominal data were learned from the model to identify samples distinct from the nominal distribution as anomalies.…”

Section: General Anomaly Detectionmentioning

confidence: 99%

N-pad : Neighboring Pixel-based Industrial Anomaly Detection

Hwang¹,

Park²

2022

Preprint

View full text Add to dashboard Cite

Identifying defects in the images of industrial products has been an important task to enhance quality control and reduce maintenance costs. In recent studies, industrial anomaly detection models were developed using pre-trained networks to learn nominal representations. To employ the relative positional information of each pixel, we present Npad, a novel method for anomaly detection and segmentation in a one-class learning setting that includes the neighborhood of the target pixel for model training and evaluation. Within the model architecture, pixel-wise nominal distributions are estimated by using the features of neighboring pixels with the target pixel to allow possible marginal misalignment. Moreover, the centroids from clusters of nominal features are identified as a representative nominal set. Accordingly, anomaly scores are inferred based on the Mahalanobis distances and Euclidean distances between the target pixel and the estimated distributions or the centroid set, respectively. Thus, we have achieved state-of-the-art performance in MVTec-AD with AUROC of 99.37 for anomaly detection and 98.75 for anomaly segmentation, reducing the error by 34% compared to the next best performing model. Experiments in various settings further validate our model.

show abstract

“…Teacher-student-based methods [4], [33], [28], [19], [10], [30] have been extensively studied in the field of AD, particularly focused on 2D images. [4] proposed a teacherstudent framework for unsupervised anomaly detection with discriminative latent embeddings to efficiently construct a descriptive teacher network.…”

Section: B Anomaly Detection With Teacher-student Networkmentioning

confidence: 99%

“…[19] proposed an adaptive framework to understand scene dynamics from audio-visual data in a hybrid fusion manner. [10] proposed a reserved distillation method by using a one-class bottleneck embedding (OCBE) module to boost the discrimination capability. Overall, a crucial problem for these methods is that a large number of normal samples are required for training, yet it is not always feasible to acquire sufficient training data.…”

Section: B Anomaly Detection With Teacher-student Networkmentioning

confidence: 99%

See 1 more Smart Citation

Teacher-Student Network for 3D Point Cloud Anomaly Detection with Few Normal Samples

Qin¹,

Gu²,

Wang³

et al. 2022

Preprint

View full text Add to dashboard Cite

Anomaly detection, which is a critical and popular topic in computer vision, aims to detect anomalous samples that are different from the normal (i.e., non-anomalous) ones. The current mainstream methods focus on anomaly detection for images, whereas little attention has been paid to 3D point cloud. In this paper, drawing inspiration from the knowledge transfer ability of teacher-student architecture and the impressive feature extraction capability of recent neural networks, we design a teacher-student structured model for 3D anomaly detection. Specifically, we use feature space alignment, dimension zoom, and max pooling to extract the features of the point cloud and then minimize a multi-scale loss between the feature vectors produced by the teacher and the student networks. Moreover, our method only requires very few normal samples to train the student network due to the teacher-student distillation mechanism. Once trained, the teacher-student network pair can be leveraged jointly to fulfill 3D point cloud anomaly detection based on the calculated anomaly score. For evaluation, we compare our method against the reconstruction-based method on the ShapeNet-Part dataset. The experimental results and ablation studies quantitatively and qualitatively confirm that our model can achieve higher performance compared with the state of the arts in 3D anomaly detection with very few training samples.

show abstract

Anomaly Detection via Reverse Distillation from One-Class Embedding

Cited by 293 publications

References 29 publications

Deep Industrial Image Anomaly Detection: A Survey

Deep Industrial Image Anomaly Detection: A Survey

N-pad : Neighboring Pixel-based Industrial Anomaly Detection

Teacher-Student Network for 3D Point Cloud Anomaly Detection with Few Normal Samples

Contact Info

Product

Resources

About