Análise de redes GANs para detecção de anomalias em atividade sonoras
Resumo
Os trabalhos do estado-da-arte na identificação de anomalias em imagens utilizam arquiteturas baseadas em GAN (Generative Adversarial Network), entretanto, poucos estudos demonstram sua utilização no domínio de sons. Testes utilizando bases de dados reais mostram que algumas alterações nas arquiteturas utilizadas para imagens podem obter resultados promissores. Validamos nossa abordagem no conjunto de dados DCASE 2020, que inclui mais de 180 horas de maquinário industrial. Avaliamos a classificação das anomalias, relatando uma média de 72% de AUC e 69% de pAUC, resultados superiores ao apresentado por baselines.
Referências
Akcay, S., Atapour Abarghouei, A., and Breckon, T. (2019a). Skip-ganomaly: Skip connected and adversarially trained encoder-decoder anomaly detection. pages 1–8.
Akcay, S., Atapour-Abarghouei, A., and Breckon, T. P. (2019b). GANomaly: Semi-supervised Anomaly Detection via Adversarial Training. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), volume 11363 LNCS, pages 622–637. Springer Verlag.
Cheng, Z., Zhu, E., Wang, S., Zhang, P., and Li, W. (2021). Unsupervised outlier detection via transformation invariant autoencoder. IEEE Access, 9:43991–44002.
Creswell, A., White, T., Dumoulin, V., Arulkumaran, K., Sengupta, B., and Bharath, A. A. (2017). Generative adversarial networks: An overview. CoRR, abs/1710.07035.
Giri, R., Tenneti, S. V., Helwani, K., Cheng, F., Isik, U., and Krishnaswamy, A. (2020). Unsupervised anomalous sound detection using self-supervised classification and group masked autoencoder for density estimation. Technical report, DCASE2020 Challenge.
Kittler, J., Kaloskampis, I., Zor, C., Xu, Y., Hicks, Y., and Wang, W. (2018). Intelligent signal processing mechanisms for nuanced anomaly detection in action audio-visual data streams. In 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 6563–6567.
Koizumi, Y., Kawaguchi, Y., and Imoto, K. (2020). Description and discussion on dcase2020 challenge task2: unsupervised anomalous sound detection for machine condition monitoring. Technical report, DCASE2020 Challenge.
Koizumi, Y., Saito, S., Uematsu, H., Harada, N., and Imoto, K. (2019). ToyADMOS: A dataset of miniature-machine operating sounds for anomalous sound detection. In Proceedings of IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA), pages 308–312.
Liu, G., Lan, S., Zhang, T., Huang, W., and Wang, W. (2021). Sagan: Skip-attention gan for anomaly detection. In 2021 IEEE International Conference on Image Processing (ICIP), pages 2468–2472.
Müller, R., Illium, S., and Linnhoff-Popien, C. (2021). Deep recurrent interpolation networks for anomalous sound detection. In 2021 International Joint Conference on Neural Networks (IJCNN), pages 1–7.
Purohit, H., Tanabe, R., Ichige, T., Endo, T., Nikaido, Y., Suefusa, K., and Kawaguchi, Y. (2019). MIMII Dataset: Sound dataset for malfunctioning industrial machine investigation and inspection. In Proceedings of the Detection and Classification of Acoustic Scenes and Events 2019 Workshop (DCASE2019), pages 209–213.
Ronneberger, O., Fischer, P., and Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. CoRR, abs/1505.04597.
Rovetta, S., Mnasri, Z., and Masulli, F. (2020). Detection of hazardous road events from audio streams: An ensemble outlier detection approach. In 2020 IEEE Conference on Evolving and Adaptive Intelligent Systems (EAIS), pages 1–6.
Schlegl, T., Seeböck, P., Waldstein, S. M., Langs, G., and Schmidt-Erfurth, U. (2019). f-anogan: Fast unsupervised anomaly detection with generative adversarial networks. Medical Image Analysis, 54:30 – 44.
Suefusa, K., Nishida, T., Purohit, H., Tanabe, R., Endo, T., and Kawaguchi, Y. (2020). Anomalous sound detection based on interpolation deep neural network. In ICASSP 2020 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 271–275.
Zenati, H., Foo, C. S., Lecouat, B., Manek, G., and Chandrasekhar, V. R. (2018). Efficient gan-based anomaly detection.
Zhou, X., Xiong, J., Zhang, X., Liu, X., and Wei, J. (2021). A radio anomaly detection algorithm based on modified generative adversarial network. IEEE Wireless Communications Letters, 10(7):1552–1556.
Akcay, S., Atapour-Abarghouei, A., and Breckon, T. P. (2019b). GANomaly: Semi-supervised Anomaly Detection via Adversarial Training. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), volume 11363 LNCS, pages 622–637. Springer Verlag.
Cheng, Z., Zhu, E., Wang, S., Zhang, P., and Li, W. (2021). Unsupervised outlier detection via transformation invariant autoencoder. IEEE Access, 9:43991–44002.
Creswell, A., White, T., Dumoulin, V., Arulkumaran, K., Sengupta, B., and Bharath, A. A. (2017). Generative adversarial networks: An overview. CoRR, abs/1710.07035.
Giri, R., Tenneti, S. V., Helwani, K., Cheng, F., Isik, U., and Krishnaswamy, A. (2020). Unsupervised anomalous sound detection using self-supervised classification and group masked autoencoder for density estimation. Technical report, DCASE2020 Challenge.
Kittler, J., Kaloskampis, I., Zor, C., Xu, Y., Hicks, Y., and Wang, W. (2018). Intelligent signal processing mechanisms for nuanced anomaly detection in action audio-visual data streams. In 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 6563–6567.
Koizumi, Y., Kawaguchi, Y., and Imoto, K. (2020). Description and discussion on dcase2020 challenge task2: unsupervised anomalous sound detection for machine condition monitoring. Technical report, DCASE2020 Challenge.
Koizumi, Y., Saito, S., Uematsu, H., Harada, N., and Imoto, K. (2019). ToyADMOS: A dataset of miniature-machine operating sounds for anomalous sound detection. In Proceedings of IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA), pages 308–312.
Liu, G., Lan, S., Zhang, T., Huang, W., and Wang, W. (2021). Sagan: Skip-attention gan for anomaly detection. In 2021 IEEE International Conference on Image Processing (ICIP), pages 2468–2472.
Müller, R., Illium, S., and Linnhoff-Popien, C. (2021). Deep recurrent interpolation networks for anomalous sound detection. In 2021 International Joint Conference on Neural Networks (IJCNN), pages 1–7.
Purohit, H., Tanabe, R., Ichige, T., Endo, T., Nikaido, Y., Suefusa, K., and Kawaguchi, Y. (2019). MIMII Dataset: Sound dataset for malfunctioning industrial machine investigation and inspection. In Proceedings of the Detection and Classification of Acoustic Scenes and Events 2019 Workshop (DCASE2019), pages 209–213.
Ronneberger, O., Fischer, P., and Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. CoRR, abs/1505.04597.
Rovetta, S., Mnasri, Z., and Masulli, F. (2020). Detection of hazardous road events from audio streams: An ensemble outlier detection approach. In 2020 IEEE Conference on Evolving and Adaptive Intelligent Systems (EAIS), pages 1–6.
Schlegl, T., Seeböck, P., Waldstein, S. M., Langs, G., and Schmidt-Erfurth, U. (2019). f-anogan: Fast unsupervised anomaly detection with generative adversarial networks. Medical Image Analysis, 54:30 – 44.
Suefusa, K., Nishida, T., Purohit, H., Tanabe, R., Endo, T., and Kawaguchi, Y. (2020). Anomalous sound detection based on interpolation deep neural network. In ICASSP 2020 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 271–275.
Zenati, H., Foo, C. S., Lecouat, B., Manek, G., and Chandrasekhar, V. R. (2018). Efficient gan-based anomaly detection.
Zhou, X., Xiong, J., Zhang, X., Liu, X., and Wei, J. (2021). A radio anomaly detection algorithm based on modified generative adversarial network. IEEE Wireless Communications Letters, 10(7):1552–1556.
Publicado
06/08/2023
Como Citar
OLIVEIRA NETO, Wilson A. de; FIGUEIREDO, Carlos Maurício S..
Análise de redes GANs para detecção de anomalias em atividade sonoras. In: SIMPÓSIO BRASILEIRO DE COMPUTAÇÃO UBÍQUA E PERVASIVA (SBCUP), 15. , 2023, João Pessoa/PB.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2023
.
p. 11-20.
ISSN 2595-6183.
DOI: https://doi.org/10.5753/sbcup.2023.230034.
