Abstract
Action recognition is a crucial task in computer vision. The Two-Stream Network and 3D ConvNets have achieved outstanding performance. However, due to the enormous amount of computation of optical flow and 3D convolution, they can not be effectively applied to some real-time applications. Therefore, some researchers have used motion vectors and residuals in the compressed video to replace optical flow, but their lack of fine structure leads to decreased model performance, such as noise and inaccurate motion blocks. In this paper, we propose a Multi-Knowledge Attention Transfer (MKAT) framework based on the three-stream network structure, including Multi-Knowledge Enhancement (MKE) module and Feature Loss Enhancement (FLE) module. The MKE module adopts the distillation methods of self-learning and multi-level information fusion, which allow the student network to learn from the decision-making and thinking aspects of the teacher. We use the attention enhancement (AE) module to process the output of each feature layer, so that the FLE module can highlight the differences between the key information of the feature layer. Experimental results on two public benchmarks (i.e., UCF-101, HMDB-51) significantly outperform the current state of the art on the compressed domain.
This work is supported by the CERNET Innovation Project No. NGII20190625, the Inner Mongolia Natural Science Foundation of China under Grant No. 2021MS06016, and the Inner Mongolia University Postgraduate Research Innovative Project No. 11200-121024.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Ba, L.J., Caruana, R.: Do deep nets really need to be deep? arXiv preprint arXiv:1312.6184 (2013)
Battash, B., Barad, H., Tang, H., Bleiweiss, A.: Mimic the raw domain: accelerating action recognition in the compressed domain. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops, pp. 684–685 (2020)
Carreira, J., Noland, E., Hillier, C., Zisserman, A.: A short note on the kinetics-700 human action dataset. arXiv preprint arXiv:1907.06987 (2019)
Carreira, J., Zisserman, A.: Quo vadis, action recognition? a new model and the kinetics dataset. In: proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 6299–6308 (2017)
Christoph, R., Pinz, F.A.: Spatiotemporal residual networks for video action recognition. Adv. Neural Inf. Process. Syst. 3468–3476 (2016)
Diba, A., et al.: Temporal 3d convnets: new architecture and transfer learning for video classification. arXiv preprint arXiv:1711.08200 (2017)
Donahue, J., et al.: Long-term recurrent convolutional networks for visual recognition and description. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 2625–2634 (2015)
Feichtenhofer, C., Pinz, A., Wildes, R.P.: Spatiotemporal multiplier networks for video action recognition. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 4768–4777 (2017)
Feichtenhofer, C., Pinz, A., Zisserman, A.: Convolutional two-stream network fusion for video action recognition. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 1933–1941 (2016)
He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recognition. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 770–778 (2016)
Hinton, G., Vinyals, O., Dean, J.: Distilling the knowledge in a neural network. arXiv preprint arXiv:1503.02531 (2015)
Huang, S., Zhao, X., Niu, L., Zhang, L.: Static image action recognition with hallucinated fine-grained motion information. In: 2021 IEEE International Conference on Multimedia and Expo (ICME), pp. 1–6. IEEE (2021)
Huang, Z., Wang, N.: Like what you like: Knowledge distill via neuron selectivity transfer. arXiv preprint arXiv:1707.01219 (2017)
Huo, Y., Xu, X., Lu, Y., Niu, Y., Lu, Z., Wen, J.R.: Mobile video action recognition. arXiv preprint arXiv:1908.10155 (2019)
Kingma, D.P., Ba, J.: Adam: a method for stochastic optimization. arXiv preprint arXiv:1412.6980 (2014)
Krizhevsky, A., Sutskever, I., Hinton, G.E.: Imagenet classification with deep convolutional neural networks. Adv. Neural. Inf. Process. Syst. 25, 1097–1105 (2012)
Kuehne, H., Jhuang, H., Garrote, E., Poggio, T., Serre, T.: HMDB: a large video database for human motion recognition. In: 2011 International Conference on Computer Vision, pp. 2556–2563. IEEE (2011)
Lan, Z., Zhu, Y., Hauptmann, A.G., Newsam, S.: Deep local video feature for action recognition. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops, pp. 1–7 (2017)
Li, B., Kong, L., Zhang, D., Bao, X., Huang, D., Wang, Y.: Towards practical compressed video action recognition: a temporal enhanced multi-stream network. In: 2020 25th International Conference on Pattern Recognition (ICPR), pp. 3744–3750. IEEE (2021)
Lin, J., Gan, C., Han, S.: Temporal shift module for efficient video understanding. 2019 IEEE. In: CVF International Conference on Computer Vision (ICCV), pp. 7082–7092 (2019)
Qiu, Z., Yao, T., Mei, T.: Learning spatio-temporal representation with pseudo-3d residual networks. In: proceedings of the IEEE International Conference on Computer Vision, pp. 5533–5541 (2017)
Shou, Z., et al.: Dmc-net: generating discriminative motion cues for fast compressed video action recognition. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 1268–1277 (2019)
Simonyan, K., Zisserman, A.: Two-stream convolutional networks for action recognition in videos. arXiv preprint arXiv:1406.2199 (2014)
Soomro, K., Zamir, A.R., Shah, M.: Ucf101: a dataset of 101 human actions classes from videos in the wild. arXiv preprint arXiv:1212.0402 (2012)
Tran, D., Bourdev, L., Fergus, R., Torresani, L., Paluri, M.: Learning spatiotemporal features with 3d convolutional networks. In: Proceedings of the IEEE International Conference on Computer Vision, pp. 4489–4497 (2015)
Tran, D., Ray, J., Shou, Z., Chang, S.F., Paluri, M.: Convnet architecture search for spatiotemporal feature learning. arXiv preprint arXiv:1708.05038 (2017)
Wang, H., Schmid, C.: Action recognition with improved trajectories. In: Proceedings of the IEEE International Conference on Computer Vision, pp. 3551–3558 (2013)
Wang, K., Gao, X., Zhao, Y., Li, X., Dou, D., Xu, C.Z.: Pay attention to features, transfer learn faster CNNs. In: International Conference on Learning Representations (2019)
Wang, L., et al.: Temporal segment networks: towards good practices for deep action recognition. In: European Conference on Computer Vision (2016)
Wu, C.Y., Zaheer, M., Hu, H., Manmatha, R., Smola, A.J., Krähenbühl, P.: Compressed video action recognition. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 6026–6035 (2018)
Wu, M.C., Chiu, C.T.: Multi-teacher knowledge distillation for compressed video action recognition based on deep learning. J. Syst. Archit. 103, 101695 (2020)
Yang, K., Qiao, P., Li, D., Dou, Y.: If-TTN: Information fused temporal transformation network for video action recognition. arXiv preprint arXiv:1902.09928 (2019)
Yue-Hei Ng, J., Hausknecht, M., Vijayanarasimhan, S., Vinyals, O., Monga, R., Toderici, G.: Beyond short snippets: deep networks for video classification. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 4694–4702 (2015)
Zach, C., Pock, T., Bischof, H.: A duality based approach for realtime TV-L1 optical flow. In: Hamprecht, F.A., Schnörr, C., Jähne, B. (eds.) DAGM 2007. LNCS, vol. 4713, pp. 214–223. Springer, Heidelberg (2007). https://doi.org/10.1007/978-3-540-74936-3_22
Zagoruyko, S., Komodakis, N.: Paying more attention to attention: improving the performance of convolutional neural networks via attention transfer. arXiv preprint arXiv:1612.03928 (2016)
Zhang, B., Wang, L., Wang, Z., Qiao, Y., Wang, H.: Real-time action recognition with enhanced motion vector CNNs. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 2718–2726 (2016)
Zhang, B., Wang, L., Wang, Z., Qiao, Y., Wang, H.: Real-time action recognition with deeply transferred motion vector CNNs. IEEE Trans. Image Process. 27(5), 2326–2339 (2018)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Zhang, J., Guo, J., Ma, M. (2022). Multi-Knowledge Attention Transfer Framework for Action Recognition. In: Pimenidis, E., Angelov, P., Jayne, C., Papaleonidas, A., Aydin, M. (eds) Artificial Neural Networks and Machine Learning – ICANN 2022. ICANN 2022. Lecture Notes in Computer Science, vol 13529. Springer, Cham. https://doi.org/10.1007/978-3-031-15919-0_20
Download citation
DOI: https://doi.org/10.1007/978-3-031-15919-0_20
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-15918-3
Online ISBN: 978-3-031-15919-0
eBook Packages: Computer ScienceComputer Science (R0)