PANDA: Adapting Pretrained Features for Anomaly Detection and Segmentation

Reiss, Tal; Cohen, Niv; Bergman, Liron; Hoshen, Yedid

doi:10.1109/cvpr46437.2021.00283

Cited by 143 publications

(89 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A different class of methods leverage descriptors from pretrained networks for anomaly detection (Bergmann et al, 2020;Cohen and Hoshen, 2020;Defard et al, 2021;Gudovskiy et al, 2022;Mishra et al, 2020;Reiss et al, 2021;Rippel et al, 2021). The key idea behind these approaches is that anomalous regions produce descriptors that differ from the ones without anomalies.…”

Section: Anomaly Detection In 2dmentioning

confidence: 99%

Anomaly Detection in 3D Point Clouds using Deep Geometric Descriptors

Bergmann¹,

Sattlegger²

2022

Preprint

View full text Add to dashboard Cite

We present a new method for the unsupervised detection of geometric anomalies in highresolution 3D point clouds. In particular, we propose an adaptation of the established student-teacher anomaly detection framework to three dimensions. A student network is trained to match the output of a pretrained teacher network on anomaly-free point clouds. When applied to test data, regression errors between the teacher and the student allow reliable localization of anomalous structures. To construct an expressive teacher network that extracts dense local geometric descriptors, we introduce a novel self-supervised pretraining strategy. The teacher is trained by reconstructing local receptive fields and does not require annotations. Extensive experiments on the comprehensive MVTec 3D Anomaly Detection dataset highlight the effectiveness of our approach, which outperforms the next-best method by a large margin. Ablation studies show that our approach meets the requirements of practical applications regarding performance, runtime, and memory consumption.

show abstract

Section: Anomaly Detection In 2dmentioning

confidence: 99%

Anomaly Detection in 3D Point Clouds using Deep Geometric Descriptors

Bergmann¹,

Sattlegger²

2022

Preprint

View full text Add to dashboard Cite

show abstract

Section: Related Workmentioning

confidence: 99%

“…Recently, networks pre-trained on the large dataset are proven to be capable of extracting discriminative features for anomaly detection [7,8,22,24,27,28]. With a pre-trained model, memorizing its anomaly-free features helps to identify anomalous samples [7,27]. The studies in [8,28] show that using the Mahalanobis distance to measure the similarity between anomalies and anomaly-free features leads to accurate anomaly detection.…”

Section: Related Workmentioning

confidence: 99%

Anomaly Detection via Reverse Distillation from One-Class Embedding

Deng¹,

Li²

2022

Preprint

View full text Add to dashboard Cite

Knowledge distillation (KD) achieves promising results on the challenging problem of unsupervised anomaly detection (AD). The representation discrepancy of anomalies in the teacher-student (T-S) model provides essential evidence for AD. However, using similar or identical architectures to build the teacher and student models in previous studies hinders the diversity of anomalous representations. To tackle this problem, we propose a novel T-S model consisting of a teacher encoder and a student decoder and introduce a simple yet effective "reverse distillation" paradigm accordingly. Instead of receiving raw images directly, the student network takes teacher model's one-class embedding as input and targets to restore the teacher's multiscale representations. Inherently, knowledge distillation in this study starts from abstract, high-level presentations to low-level features. In addition, we introduce a trainable one-class bottleneck embedding (OCBE) module in our T-S model. The obtained compact embedding effectively preserves essential information on normal patterns, but abandons anomaly perturbations. Extensive experimentation on AD and one-class novelty detection benchmarks shows that our method surpasses SOTA performance, demonstrating our proposed approach's effectiveness and generalizability.

show abstract

“…We use the official code by the respective authors 34 . NTL is built upon the final pooling layer of a pre-trained ResNet152 for CIFAR-10 and F-MNIST (as suggested in Defard et al ( 2021)), and upon the third residual block of a pre-trained WideResNet50 for MVTEC (as suggested in Reiss et al (2021)). Further implementation details of NTL are in the Appendix C. We adopt the two proposed LOE methods (Section 3) and the two baseline methods "Blind" and "Refine" (Section 2) to both backbone models.…”

Section: Experiments On Image Datamentioning

confidence: 99%

“…NTL on image data NTL is built upon the final pooling layer of a pre-trained ResNet152 on CIFAR-10 and F-MNIST (as suggested in Defard et al ( 2021)), and upon the third residual block of a pre-trained WideResNet50 on MVTEC (as suggested in Reiss et al (2021)). On all image datasets, the pre-trained feature extractors are frozen during training.…”

Section: Implementation Detailsmentioning

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

PANDA: Adapting Pretrained Features for Anomaly Detection and Segmentation

Cited by 143 publications

References 9 publications

Anomaly Detection in 3D Point Clouds using Deep Geometric Descriptors

Anomaly Detection in 3D Point Clouds using Deep Geometric Descriptors

Anomaly Detection via Reverse Distillation from One-Class Embedding

Latent Outlier Exposure for Anomaly Detection with Contaminated Data

Contact Info

Product

Resources

About