Asymmetric Student-Teacher Networks for Industrial Anomaly Detection

Rudolph, Marco; Wehrbein, Tom; Rosenhahn, Bodo; Wandt, Bastian

doi:10.1109/wacv56688.2023.00262

Cited by 80 publications

(31 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In recent years the field has put significant effort in constructing bijectors that conform with these restrictions but remain sufficiently expressive and computationally efficient even in high-dimensional problems such as images (Louizos & Welling 2017;Rothfuss et al 2019;Nielsen et al 2020;Wu et al 2020;Zhang & Chen 2021) and notable bijector architectures including MADE (Germain et al 2015), Masked Autoregressive Flow (Papamakarios et al 2017), NICE (Dinh et al 2014), RealNVP (Dinh et al 2016), Sylvester NF (Berg et al 2018), FFJORD (Grathwohl et al 2018), Glow (Kingma & Dhariwal 2018), and NSF (Durkan et al 2019). NF has proven to be a highly successful approach in a wide range of applications, including audio synthesis (Oord et al 2018;Prenger et al 2019;Aggarwal et al 2020), text translation (Jin et al 2019;Izmailov et al 2020), anomaly detection (Rudolph et al 2021;Gudovskiy et al 2022), time series forcasting (Schmidt & Simic 2019;Rasul et al 2020;Feng et al 2022), and image generation (Grathwohl et al 2018;Kingma & Dhariwal 2018;Lugmayr et al 2020).…”

Section: Nfsmentioning

confidence: 99%

To Sample or Not to Sample: Retrieving Exoplanetary Spectra with Variational Inference and Normalizing Flows

Yip,

Changeat,

Al-Refaie

et al. 2024

ApJ

View full text Add to dashboard Cite

Current endeavours in exoplanet characterization rely on atmospheric retrieval to quantify crucial physical properties of remote exoplanets from observations. However, the scalability and efficiency of said technique are under strain with increasing spectroscopic resolution and forward model complexity. The situation has become more acute with the recent launch of the James Webb Space Telescope and other upcoming missions. Recent advances in machine learning provide optimization-based variational inference as an alternative approach to perform approximate Bayesian posterior inference. In this investigation we developed a normalizing-flow-based neural network, combined with our newly developed differentiable forward model, Diff-τ, to perform Bayesian inference in the context of atmospheric retrievals. Using examples from real and simulated spectroscopic data, we demonstrate the advantages of our proposed framework: (1) training our neural network does not require a large precomputed training set and can be trained with only a single observation; (2) it produces high-fidelity posterior distributions in excellent agreement with sampling-based retrievals; (3) it requires up to 75% fewer forward model calls to converge to the same result; and (4) this approach allows formal Bayesian model selection. We discuss the computational efficiencies of Diff-τ in relation to TauREx3's nominal forward model and provide a “lessons learned” account of developing radiative transfer models in differentiable languages. Our proposed framework contributes toward the latest development of neural network–powered atmospheric retrieval. Its flexibility and significant reduction in forward model calls required for convergence holds the potential to be an important addition to the retrieval tool box for large and complex data sets along with sampling-based approaches.

show abstract

Section: Nfsmentioning

confidence: 99%

To Sample or Not to Sample: Retrieving Exoplanetary Spectra with Variational Inference and Normalizing Flows

Yip,

Changeat,

Al-Refaie

et al. 2024

ApJ

View full text Add to dashboard Cite

show abstract

“…IKD [15] Context Similarity ResNet The paper adds context similarity loss and adaptive hard sample mining module to prevent overfitting. AST [16] L2, Log-Likelihood EfficientNet The paper uses a asymmetric teacher-student network to make the representation of anomaly more different.…”

Section: Resnetmentioning

confidence: 99%

“…The abnormal image features extracted by the teacher-student network of RD4AD differ significantly during inference. AST [16] concludes that the abnormal image features extracted by the teacher-student model with the same structure are significantly similar, so they propose an asymmetric teacher-student architecture to address this issue. AST also introduces a normalized flow to avoid this problem and prevent estimation bias caused by the inconsistency of the two network structures.…”

Section: Resnetmentioning

confidence: 99%

See 1 more Smart Citation

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

“…Therefore, detecting point clouds with structural anomalies remains challenging. Few methods [21,28,29] are available for 3D industrial point cloud defect detection due to the scarcity of 3D industrial point cloud datasets. However, 3D point clouds represent the surface geometry information of the target industrial products, which is not available in 2D images.…”

Section: Introductionmentioning

confidence: 99%

Defect segmentation with local embedding in industrial 3D point clouds based on transformer

Jing,

Wang

2023

Meas. Sci. Technol.

View full text Add to dashboard Cite

Three-dimensional (3D) defect detection provides an effective method for improving industrial production efficiency. However, the 3D dataset is scarce, which is valuable for the industrial production field. This study proposes a new approach for detecting defect point clouds, which can provide an end-to-end 3D defect detection model. A self-attention mechanism is used to enrich the semantic relationships between local neighborhood features and global features based on the connection between them. Through adding multi-channel features, the rich structural features of the target point cloud are obtained, and the defect areas are accurately segmented to finally complete the 3D point cloud defect detection task. Furthermore, the multi-feature fusion in the model makes the segmented defect regions closer to the ground truth. Our method outperforms four state-of-the-art point cloud segmentation methods in terms of both segmentation region accuracy and defect detection point cloud accuracy. In the field of 3D defect detection, it provides an effective method to detect 3D information of industrial products.

show abstract

Asymmetric Student-Teacher Networks for Industrial Anomaly Detection

Cited by 80 publications

References 39 publications

To Sample or Not to Sample: Retrieving Exoplanetary Spectra with Variational Inference and Normalizing Flows

To Sample or Not to Sample: Retrieving Exoplanetary Spectra with Variational Inference and Normalizing Flows

Deep Industrial Image Anomaly Detection: A Survey

Defect segmentation with local embedding in industrial 3D point clouds based on transformer

Contact Info

Product

Resources

About