Abstract
Novelty detection (ND) methods seek to identify anomalies within a specific dataset. Although self-supervised representation learning is commonly used in such applications, inadequate training data may reduce the effectiveness of these methods. It is thus reasonable to use external data to improve these performances. Here, we propose a simple and effective network, CLOE, for image-based novelty detection. Our method includes a pretrained ViT model as a feature extractor and employs the contrastive learning technique to train the dataset with external data. We compare the performance of two types of extra training settings: (1) The augmented data of the original dataset. (2) The fake images obtained from generative models. The demonstrated approach achieves a new state-of-the-art performance in novelty detection, as evidenced by achieving an ROC-AUC of 99.72% on the CIFAR-10 dataset.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Salehi, M., Mirzaei, H., Hendrycks, D., Li, Y., Rohban, M.H., Sabokrou, M.: A unified survey on anomaly, novelty, open-set, and out-of-distribution detection: solutions and future challenges. ar**v preprint ar**v:2110.14051 (2021)
Dosovitskiy, A., et al.: An image is worth 16\(\times \)16 words: transformers for image recognition at scale (2021)
Reiss, T., Cohen, N., Bergman, L., Hoshen, Y.: Panda: adapting pretrained features for anomaly detection and segmentation. In: 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 2805–2813. IEEE (2021)
Hendrycks, D., Mazeika, M., Dietterich, T.: Deep anomaly detection with outlier exposure (2018)
Chen, T., Kornblith, S., Norouzi, M., Hinton, G.: A simple framework for contrastive learning of visual representations. In: International conference on machine learning, pp. 1597–1607. PMLR (2020)
Krizhevsky, A., Hinton, G., et al.: Learning multiple layers of features from tiny images (2009)
Bergman, L., Cohen, N., Hoshen, Y.: Deep nearest neighbor anomaly detection. ar**v preprint ar**v:2002.10445 (2020)
Ruff, L., Vandermeulen, R.A., Görnitz, N., Deecke, L., Kloft, M.: Deep one-class classification. In: International Conference on Machine Learning (2018)
Salehi, M., Sadjadi, N., Baselizadeh, S., Rohban, M.H., Rabiee, H.R.: Multiresolution knowledge distillation for anomaly detection. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 14902–14912 (2021)
Liu, Z., Zhou, Y., Xu, Y., Wang, Z.: SimpleNet: a simple network for image anomaly detection and localization. ar**v preprint ar**v:2303.15140 (2023)
Murase, H., Fukumizu, K.: ALGAN: anomaly detection by generating pseudo anomalous data via latent variables. IEEE Access 10, 44259–44270 (2022)
Cohen, M.J., Avidan, S.: Transformaly-two (feature spaces) are better than one. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 4060–4069 (2022)
Mirzaei, H., et al.: Fake it until you make it: towards accurate near-distribution novelty detection. In: NeurIPS ML Safety Workshop
Heusel, M., Ramsauer, H., Unterthiner, T., Nessler, B., Hochreiter, S.: GANs trained by a two time-scale update rule converge to a local nash equilibrium. In: Advances in Neural Information Processing Systems, vol. 30 (2017)
Hjelm, R.D., et al.: Learning deep representations by mutual information estimation and maximization. ar**v preprint ar**v:1808.06670 (2018)
He, K., Fan, H., Wu, Y., **e, S., Girshick, R.: Momentum contrast for unsupervised visual representation learning. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 9729–9738 (2020)
Reiss, T., Hoshen, Y.: Mean-shifted contrastive loss for anomaly detection. ar**v preprint ar**v:2106.03844 (2021)
Ridnik, T., Ben-Baruch, E., Noy, A., Zelnik-Manor, L.: ImageNet-21k pretraining for the masses. ar**v preprint ar**v:2104.10972 (2021)
Hendrycks, D., Mazeika, M., Kadavath, S., Song, D.: Using self-supervised learning can improve model robustness and uncertainty. In: Advances in Neural Information Processing Systems, vol. 32 (2019)
Sohn, K., Li, C.L., Yoon, J., **, M., Pfister, T.: Learning and evaluating representations for deep one-class classification. ar**v preprint ar**v:2011.02578 (2020)
Tack, J., Mo, S., Jeong, J., Shin, J.: CSI: novelty detection via contrastive learning on distributionally shifted instances. Adv. Neural. Inf. Process. Syst. 33, 11839–11852 (2020)
Liznerski, P., Ruff, L., Vandermeulen, R.A., Franks, B.J., Müller, K.R., Kloft, M.: Exposing outlier exposure: what can be learned from few, one, and zero outlier images. ar**v preprint ar**v:2205.11474 (2022)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
About this paper
Cite this paper
Liu, T., Liang, Q., Yang, H. (2023). CLOE: Novelty Detection via Contrastive Learning with Outlier Exposure. In: Yang, H., et al. Intelligent Robotics and Applications. ICIRA 2023. Lecture Notes in Computer Science(), vol 14273. Springer, Singapore. https://doi.org/10.1007/978-981-99-6498-7_4
Download citation
DOI: https://doi.org/10.1007/978-981-99-6498-7_4
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-99-6497-0
Online ISBN: 978-981-99-6498-7
eBook Packages: Computer ScienceComputer Science (R0)