Spatio-Temporal Anomaly Detection (WSSTAD) in surveillance video.
TEMPORAL ANOMALY PDF
In: 2018 IEEE International Conference on Data Mining, pp.Download a PDF of the paper titled Weakly-Supervised Spatio-Temporal Anomaly Detection in Surveillance Video, by Jie Wu and 7 other authors Download PDF Abstract: In this paper, we introduce a novel task, referred to as Weakly-Supervised Zenati, H., Romain, M., Foo, C.S., Lecouat, B., Chandrasekhar., V.: Adversarially learned anomaly detection. Yang, S., et al.: Phenotype analysis method for identification of gene functions involved in asymmetric division of caenorhabditis elegans.
TEMPORAL ANOMALY SERIES
Ueda, T., Seo, M., Tohsato, Y., Nishikawa, I.: Analysis of time series anomalies using causal InfoGAN and its application to biological data. In: Kůrková, V., Manolopoulos, Y., Hammer, B., Iliadis, L., Maglogiannis, I. Ueda, T., Seo, M., Nishikawa, I.: Data correction by a generative model with an encoder and its application to structure design. Schlegl, T., Seeböck, P., Waldstein, S.M., Schmidt-Erfurth, U., Langs, G.: Unsupervised anomaly detection with generative adversarial networks to guide marker discovery. Schlegl, T., Seeböck, P., Waldstein, S.M., Langs, G., Schmidt-Erfurth., U.: f-AnoGAN: fast unsupervised anomaly detection with generative adversarial networks. In: International Conference on Learning Representations (2018) Miyato, T., Kataoka, T., Koyama, M., Yoshida., Y.: Spectral normalization for generative adversarial networks. Lin, Z., Thekumparampil, K.K., Fanti, G.C., Oh, S.: InfoGAN-CR: disentangling generative adversarial networks with contrastive regularizers. In: Tetko, I.V., Kůrková, V., Karpov, P., Theis, F. Li, D., Chen, D., Jin, B., Shi, L., Goh, J., Ng, S.-K.: MAD-GAN: multivariate anomaly detection for time series data with generative adversarial networks. Kyoda, K., et al.: WDDD: worm developmental dynamics database. In: Advances in Neural Information Processing Systems, pp. Kurutach, T., T.A., Yang, G., Russell, S., Abbeel, P.: Learning plannable representations with causal InfoGAN. In: International Conference on Learning Representations (2014)
Kingma, D., Welling, M.: Auto-encoding variational bayes. Karras, T., Aila, T., Laine, S., Lehtinen, J.: Progressive growing of GANs for improved quality, stability, and variation. In: Proceedings of the 32nd International Conference on Machine Learning, vol. Ioffe, S., Szegedy., C.: Batch normalization: Accelerating deep network training by reducing internal covariate shift. Guo, Y., Liao, W., Wang, Q., Yu, L., Ji, T., Li, P.: Multidimensional time series anomaly detection: a GRU-based gaussian mixture variational autoencoder approach. Goodfellow, I., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D.: Generative adversarial nets. KeywordsĬhen, X., Duan, Y., Houthooft, R., Schulman, J., Sutskever, I., Abbeel, P.: InfoGAN: interpretable representation learning by information maximizing generative adversarial nets. elegans, which lead to the detection of morphological and temporal anomalies caused by the knockdown of lethal genes. Computer experiments are conducted on three-dimensional data of the cell (nuclear) division dynamics in early embryonic development of C. The proposed method is applied to four-dimensional biological dataset: morphological data of a genetically manipulated embryo. Temporal anomalies are quantified by the transitions in the acquired state space. The present study proposes a method to characterize temporal anomalies in time series using Causal InfoGAN, proposed by Kurutach et al., to disentangle a state space of the dynamics of time-series data. If the latent space is disentangled (in a sense that some latent variables are interpretable and can characterize the data), the anomaly is also characterized by the mapped position in the latent space. Then, a data anomaly can be detected by a reconstruction error and a position in the latent space.
In this study, generative adversarial networks (GAN) are used as the normal data generator, and an additional encoder is trained to map data to the latent space. An approach to anomaly detection is to use a partly disentangled representation of the latent space of a generative model.
0 kommentar(er)
