Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
184 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
45 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

[Reproducibility Report] Explainable Deep One-Class Classification (2206.02598v2)

Published 6 Jun 2022 in cs.CV, cs.LG, and stat.ML

Abstract: Fully Convolutional Data Description (FCDD), an explainable version of the Hypersphere Classifier (HSC), directly addresses image anomaly detection (AD) and pixel-wise AD without any post-hoc explainer methods. The authors claim that FCDD achieves results comparable with the state-of-the-art in sample-wise AD on Fashion-MNIST and CIFAR-10 and exceeds the state-of-the-art on the pixel-wise task on MVTec-AD. We reproduced the main results of the paper using the author's code with minor changes and provide runtime requirements to achieve if (CPU memory, GPU memory, and training time). We propose another analysis methodology using a critical difference diagram, and further investigate the test performance of the model during the training phase.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (35)
  1. Academic torrents, download page for ImageNet22k version \sayfall 2011. https://academictorrents.com/details/564a77c1e1119da199ff32622a1609431b9f1c47, 2022.
  2. Documentation page of gpustat. https://github.com/wookayin/gpustat, 2022.
  3. Documentation page of the python library psutil. https://psutil.readthedocs.io/en/latest/, 2022.
  4. GitHub repository liznerski/fcdd. https://github.com/liznerski/fcdd, 2022. Downloaded: 2021-12-20. Forked on commit 7af3d8eadabee81ab8f7db5dea7f8389ef090213.
  5. ImageNet official page, challenge ILSVRC 2012. https://image-net.org/challenges/LSVRC/2012/2012-downloads.php, 2022.
  6. NVIDIA’s TITAN X product page. https://www.nvidia.com/en-us/geforce/products/10series/titan-x-pascal/, 2022. Accessed: 2022-01-24.
  7. NVIDIA’s TITAN Xp product page. https://www.nvidia.com/en-us/titan/titan-xp/, 2022. Accessed: 2022-01-24.
  8. Papers with code’s leaderboard on Anomaly Detection on MVTec AD. https://paperswithcode.com/sota/anomaly-detection-on-mvtec-ad?metric=Segmentation, 2022.
  9. The MVTec Anomaly Detection Dataset: A Comprehensive Real-World Dataset for Unsupervised Anomaly Detection. International Journal of Computer Vision, 129(4):1038–1059, Apr. 2021.
  10. MVTec AD – A Comprehensive Real-World Dataset for Unsupervised Anomaly Detection. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2019.
  11. EMNIST: Extending MNIST to handwritten letters. In 2017 International Joint Conference on Neural Networks (IJCNN), pages 2921–2926, 2017.
  12. Iterative energy-based projection on a normal data manifold for anomaly localization. page 17, 2020.
  13. ImageNet: A large-scale hierarchical image database. In 2009 IEEE Conference on Computer Vision and Pattern Recognition, pages 248–255, 2009.
  14. The Pascal Visual Object Classes (VOC) Challenge. International Journal of Computer Vision, 88(2):303–338, June 2010.
  15. CFLOW-AD: Real-Time Unsupervised Anomaly Detection with Localization via Conditional Normalizing Flows. arXiv:2107.12571 [cs], July 2021. arXiv: 2107.12571 version: 1.
  16. Deep Anomaly Detection with Outlier Exposure. In International Conference on Learning Representations, 2019.
  17. Deep learning for time series classification: a review. Data Mining and Knowledge Discovery, 33(4):917–963, 2019.
  18. Semi-orthogonal Embedding for Efficient Unsupervised Anomaly Segmentation. arXiv:2105.14737 [cs], May 2021. arXiv: 2105.14737 version: 1.
  19. A. Krizhevsky. Learning Multiple Layers of Features from Tiny Images, 2009. Technical report.
  20. Analyzing Classifiers: Fisher Vectors and Deep Neural Networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 2016.
  21. Superpixel Masking and Inpainting for Self-Supervised Anomaly Detection. page 12, 2020.
  22. Towards Visually Explaining Variational Autoencoders. In IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2020.
  23. Explainable Deep One-Class Classification. In International Conference on Learning Representations, 2021.
  24. Understanding the Effective Receptive Field in Deep Convolutional Neural Networks. In Proceedings of the 30th International Conference on Neural Information Processing Systems, NIPS’16, pages 4905–4913, Red Hook, NY, USA, 2016. Curran Associates Inc. event-place: Barcelona, Spain.
  25. Anomaly Detection in Nanofibrous Materials by CNN-Based Self-Similarity. Sensors (Basel, Switzerland), 18(1):209, Jan. 2018.
  26. O. Pfungst. Clever Hans (The Horse of Mr. Von Osten) A contribution to experimental animal and human psychology. Oct. 2010.
  27. Towards Total Recall in Industrial Anomaly Detection. arXiv:2106.08265 [cs], June 2021. arXiv: 2106.08265 version: 1.
  28. Deep one-class classification. In J. Dy and A. Krause, editors, Proceedings of the 35th International Conference on Machine Learning, volume 80 of Proceedings of Machine Learning Research, pages 4393–4402. PMLR, 10–15 Jul 2018.
  29. Rethinking Assumptions in Deep Anomaly Detection. arXiv:2006.00339 [cs, stat], July 2021. arXiv: 2006.00339.
  30. Unsupervised Anomaly Detection with Generative Adversarial Networks to Guide Marker Discovery. In M. Niethammer, M. Styner, S. Aylward, H. Zhu, I. Oguz, P.-T. Yap, and D. Shen, editors, Information Processing in Medical Imaging, pages 146–157, Cham, 2017. Springer International Publishing.
  31. K. A. Spackman. Signal Detection Theory: Valuable Tools for Evaluating Inductive Learning. In Proceedings of the Sixth International Workshop on Machine Learning, pages 160–163, Ithaca, New York, USA, 1989. Morgan Kaufmann Publishers Inc., 340 Pine Street, Sixth Floor, San Francisco, CA, USA.
  32. Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms, 2017. arXiv: 1708.07747.
  33. FastFlow: Unsupervised Anomaly Detection and Localization via 2D Normalizing Flows. arXiv:2111.07677 [cs], Nov. 2021. arXiv: 2111.07677 version: 2.
  34. Focus Your Distribution: Coarse-to-Fine Non-Contrastive Learning for Anomaly Detection and Localization. arXiv:2110.04538 [cs], Oct. 2021. arXiv: 2110.04538 version: 1.
  35. Encoding Structure-Texture Relation with P-Net for Anomaly Detection in Retinal Images. In A. Vedaldi, H. Bischof, T. Brox, and J.-M. Frahm, editors, Computer Vision – ECCV 2020, pages 360–377, Cham, 2020. Springer International Publishing.

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now