Modeling the Distribution of Normal Data in Pre-Trained Deep Features for Anomaly Detection

Rippel, Oliver; Mertens, Patrick; Merhof, Dorit

doi:10.1109/icpr48806.2021.9412109

Cited by 135 publications

(86 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Similarly, [14] recently proposed PaDiM, which utilizes a locally constrained bag-of-features approach [8], estimating patch-level feature distribution moments (mean and covariance) for patch-level Mahalanobis distance measures [33]. This approach is similar to [40] studied on full images. To better account for the distribution shift between natural pretraining and industrial image data, subsequent adaptation can be done, e.g.…”

Section: Related Workmentioning

confidence: 99%

Towards Total Recall in Industrial Anomaly Detection

Roth¹,

Pemula²,

Zepeda³

et al. 2021

Preprint

View full text Add to dashboard Cite

Being able to spot defective parts is a critical component in large-scale industrial manufacturing. A particular challenge that we address in this work is the cold-start problem: fit a model using nominal (non-defective) example images only. While handcrafted solutions per class are possible, the goal is to build systems that work well simultaneously on many different tasks automatically. The best peforming approaches combine embeddings from ImageNet models with an outlier detection model. In this paper, we extend on this line of work and propose PatchCore, which uses a maximally representative memory bank of nominal patchfeatures. PatchCore offers competitive inference times while achieving state-of-the-art performance for both detection and localization. On the standard dataset MVTec AD Patch-Core achieves an image-level anomaly detection AUROC score of 99.1%, more than halving the error compared to the next best competitor. We further report competitive results on two additional datasets and also find competitive results in the few samples regime. * Work done during an internship at Amazon Tübingen.1 Commonly also dubbed one-class classification (OCC).

show abstract

Section: Related Workmentioning

confidence: 99%

Towards Total Recall in Industrial Anomaly Detection

Roth¹,

Pemula²,

Zepeda³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…where the below Theorem 1 shows the optimal W is the eigenvectors related to the k-smallest eigenvalues of C i,j . Notice that 1) the computational complexity of the equation is cubically reduced to O(HW k 3 ) set aside the cost of SVD, although which is the concern, 2) PCA embedding would fail to minimize approximation error since it uses the k-largest eigenvectors [14], and 3) near-zero eigenvalues may induce substantial anomaly scores. For the last, a previous work suggests to use C + I for the inverse to avoid a possible numerical problem [7], what we follow.…”

Section: Low-rank Approximation Of Precision Matrixmentioning

confidence: 99%

Semi-orthogonal Embedding for Efficient Unsupervised Anomaly Segmentation

Kim¹,

Kim²,

Yi³

et al. 2021

Preprint

View full text Add to dashboard Cite

We present the efficiency of semi-orthogonal embedding for unsupervised anomaly segmentation. The multi-scale features from pre-trained CNNs are recently used for the localized Mahalanobis distances with significant performance. However, the increased feature size is problematic to scale up to the bigger CNNs, since it requires the batch-inverse of multi-dimensional covariance tensor. Here, we generalize an ad-hoc method, random feature selection, into semi-orthogonal embedding for robust approximation, cubically reducing the computational cost for the inverse of multi-dimensional covariance tensor. With the scrutiny of ablation studies, the proposed method achieves a new state-of-the-art with significant margins for the MVTec AD, KolektorSDD, KolektorSDD2, and mSTC datasets. The theoretical and empirical analyses offer insights and verification of our straightforward yet cost-effective approach.

show abstract

“…The anomaly score is in this case the distance between embedding vectors of a test image and reference vectors representing normality from the training dataset. The normal reference can be the center of a nsphere containing embeddings from normal images [4], [22], parameters of Gaussian distributions [23], [26] or the entire set of normal embedding vectors [5], [24]. The last option is used by SPADE [5] which has the best reported results for anomaly localization.…”

Section: Related Workmentioning

confidence: 99%

“…However, the normal class in PaDiM is described through a set of Gaussian distributions that also model correlations between semantic levels of the used pretrained CNN model. Inspired by [5], [23], we choose as pretrained networks a ResNet [27], a Wide-ResNet [28] or an EfficientNet [29]. Thanks to this modelisation, PaDiM outperforms the current state-of-the-art methods.…”

Section: Related Workmentioning

confidence: 99%

“…Inspired by [23], [26], we use the Mahalanobis distance [31] M (x ij ) to give an anomaly score to the patch in position (i, j) of a test image. M (x ij ) can be interpreted as the distance between the test patch embedding x ij and learned distribution N (µ ij , Σ ij ), where M (x ij ) is computed as follows:…”

Section: Inference : Computation Of the Anomaly Mapmentioning

confidence: 99%

See 1 more Smart Citation

PaDiM: a Patch Distribution Modeling Framework for Anomaly Detection and Localization

Defard¹,

Setkov²,

Loesch³

et al. 2020

Preprint

View full text Add to dashboard Cite

We present a new framework for Patch Distribution Modeling, PaDiM, to concurrently detect and localize anomalies in images in a one-class learning setting. PaDiM makes use of a pretrained convolutional neural network (CNN) for patch embedding, and of multivariate Gaussian distributions to get a probabilistic representation of the normal class. It also exploits correlations between the different semantic levels of CNN to better localize anomalies. PaDiM outperforms current state-ofthe-art approaches for both anomaly detection and localization on the MVTec AD and STC datasets. To match real-world visual industrial inspection, we extend the evaluation protocol to assess performance of anomaly localization algorithms on non-aligned dataset. The state-of-the-art performance and low complexity of PaDiM make it a good candidate for many industrial applications.

show abstract

Modeling the Distribution of Normal Data in Pre-Trained Deep Features for Anomaly Detection

Cited by 135 publications

References 18 publications

Towards Total Recall in Industrial Anomaly Detection

Towards Total Recall in Industrial Anomaly Detection

Semi-orthogonal Embedding for Efficient Unsupervised Anomaly Segmentation

PaDiM: a Patch Distribution Modeling Framework for Anomaly Detection and Localization

Contact Info

Product

Resources

About