臺灣博碩士論文加值系統

隨著智能城市的發展，大規模監控影像分析變得越來越重要；然而，最先進的深度學習模型所需的大量計算資源使得實時處理難以實現。由於監控視頻中普遍存在高場景相似性的特點，我們採用了一種稱為動態卷積的有效壓縮方法，它可以重複使用以前的特徵圖來減少計算量，並結合過濾器剪枝以進一步提高性能。在本論文中，我們在45個具有各種場景的監控影像上測試所提出的方法。實驗結果表示，混合剪枝結構可以減少高達80.4％的FLOPs，同時保持精度在1.34% mAP之內。此外，與基線相比，該方法可以將處理速度提高2.8倍。

The large-scale surveillance videos analysis becomes important as the development of the intelligent city; however, the heavy computational resources necessary for the state-of-the-art deep learning model makes real-time processing hard to be implemented. As the characteristic of high scene similarity generally existing in surveillance videos, we exploit an effective compression architecture called dynamic convolution, which can reuse the previous feature maps to reduce the calculation amount; and combine with filter pruning to further speed up the performance. In this paper, we tested the presented method on 45 surveillance videos with various scenes. The experimental results show that the hybrid pruning architecture can reduce up to 80.4% of FLOPs while preserving the precision within 1.34% mAP; furthermore, the method can improve the processing speed up to 2.8 times compared to the traditional Single Shot MultiBox Detection.

Chapter 1 Introduction
1.1 Background Information
1.2 Research Motivation
1.3 Research Purpose
1.4 Contributions
1.5 Organization

Chapter 2 Related Works
2.1 Quantization
2.2 Effective Model Design
2.3 Pruning

Chapter 3 Presented Approach
3.1 Traditional Convolution
3.1.1 Artificial neural network
3.1.2 Convolutional neural network
3.2 Dynamic Convolution
3.2.1 Frame differencing
3.2.2 Prediction
3.2.3 Dyn-convolution
3.3 Practical Concern and Implementation
3.3.1 Cell difference map
3.3.2 Cell-based partition
3.3.3 Index table
3.4 Presented Framework
3.5 Filter Pruning

Chapter 4 Experiments
4.1 Accuracy Performance on Datasets
4.1.1 Environment
4.1.2 On PETS 2009
4.1.3 On AVSS 2007
4.1.4 On VIRAT
4.1.5 On YouTube
4.2 Threshold of Frame Differencing
4.3 Overview Results
4.4 Effect on scene similarity

Chapter 5 Conclusions

References

Appendices
Appendix 1
Appendix 2
Appendix 3

[1] E. Ahmed, M. Jones, T. K. Marks, An improved deep learning architecture for person re-identification, IEEE Conference on Computer Vision and Pattern Recognition (2015) 3908–3916.

[2] X. Liu, W. Liu, T. Mei, H. Ma, A deep learning-based approach to progressive vehicle re-identification for urban surveillance, European Conference on Computer Vision (2016) 869–884.

[3] Y. Shen, T. Xiao, H. Li, S. Yi, X. Wang, Learning deep neural networks for vehicle re-id with visual-spatio-temporal path proposals, IEEE International Conference on Computer Vision (2017) 1900–1909.

[4] Y. Tian, P. Luo, X. Wang, X. Tang, Pedestrian detection aided by deep learning semantic tasks, IEEE Conference on Computer Vision and Pattern Recognition (2015) 5079–5087.

[5] K. Kang, X. Wang, Fully convolutional neural networks for crowd segmentation, arXiv preprint arXiv:1411.4464 (2014).

[6] BBC News, CCTV: Too many cameras useless, warns surveillance watchdog Tony Porter (2015), http://www.bbc.com/news/uk-30978995, Accessed: 2018-08-21.

[7] K. Simonyan, A. Zisserman, Very deep convolutional networks for large scale image recognition, International Conference on Learning Representations (2015).

[8] C. Szegedy, W. Liu, Y. Jia, P. Sermanet, S. E. Reed, D. Anguelov, D. Erhan, V. Vanhoucke, A. Rabinovich, Going deeper with convolutions, IEEE Conference on Computer Vision and Pattern Recognition (2015) 1–9.

[9] K. He, X. Zhang, S. Ren, J. Sun, Deep residual learning for image recogni tion, IEEE Conference on Computer Vision and Pattern Recognition (2016) 770–778.

[10] D.-Q. Gao, An Effective Dynamic Convolutional Neural Network for Surveillance Video (2018), https://hdl.handle.net/11296/2aqdrh.

[11] W. Liu, D. Anguelov, D. Erhan, C. Szegedy, S. E. Reed, C.-Y. Fu, A. C. Berg, Ssd: single shot multibox detector, European Conference on Computer Vision (2016) 21–37.

[12] S. Han, H. Mao, W. J. Dally, Deep compression: Compressing deep neural network with pruning, trained quantization and huffman coding, International Conference on Learning Representations (2016).

[13] B. Jacob, S. Kligys, B. Chen, M. Zhu, M. Tang, A. Howard, H. Adam, D. Kalenichenko, Quantization and training of neural networks for efficient integer-arithmetic-only inference, IEEE/CVF Conference on Computer Vision and Pattern Recognition (2018) 2704–2713.

[14] S. Han, X. Liu, H. Mao, J. Pu, A. Pedram, M. Horowitz, W. J. Dally, Eie: Efficient inference engine on compressed deep neural network, Annual ACM/IEEE International Symposium on Computer Architecture (2016) 243–254.

[15] Y. Chen, T. Luo, S. Liu, S. Zhang, L. He, J. Wang, L. Li, T. Chen, Z. Xu, N. Sun, O. Temam, Dadiannao: A machine-learning supercomputer, Annual IEEE/ACM International Symposium on Microarchitecture (2014) 609–622.

[16] J. Wu, C. Leng, Y. Wang, Q. Hu, J. Cheng, Quantization and training of neural networks for efficient integer-arithmetic-only inference, IEEE Conference on Computer Vision and Pattern Recognition (2016) 4820–4828.

[17] M. Rastegari, V. Ordonez, J. Redmon, A. Farhadi, Xnor net: Imagenet classification using binary convolutional neural networks, European Conference on Computer Vision (2016) 525–542.

[18] A. G. Howard, M. Zhu, B. Chen, D. Kalenichenko, W. Wang, T. W. M. Andreetto, H. Adam, Mobilenets: Efficient convolutional neural networks for mobile vision applications, arXiv preprint arXiv:1704.04861 (2017).

[19] X. Zhang, X. Zhou, M. Lin, J. Sun, Shufflenet: An extremely efficient convolutional neural network for mobile devices, IEEE/CVF Conference on Computer Vision and Pattern Recognition (2018) 6848–6856.

[20] C. Szegedy, V. Vanhoucke, S. Ioffe, J. Shlens, Z. Wojna, Rethinking the inception architecture for computer vision, IEEE conference on Computer Vision and Pattern Recognition (2016) 2818–2826.

[21] C. Szegedy, S. Ioffe, V. Vanhoucke, Inception-v4, inception-resnet and the impact of residual connections on learning, AAAI Conference on Artificial Intelligence (2017).

[22] Y.-D. Kim, E. Park, S. Yoo, T. Choi, L. Yang, D. Shin, Compression of deep convolutional neural networks for fast and low power mobile applications, International Conference on Learning Representations (2016).

[23] F. N. Iandola, M. W. Moskewicz, K. Ashraf, S. Han, W. J. Dally, K. Keutzer, Squeezenet: Alexnet-level accuracy with 50x fewer parameters and <0.5mb model size, arXiv preprint arXiv:1602.07360 (2016).

[24] S. Anwar, K. Hwang, W. Sung, Structured pruning of deep convolutional neural networks, ACM Journal on Emerging Technologies in Computing Systems 13 (3) (2017) 32.

[25] A. Polyak, L. Wolf, Channel-level acceleration of deep face representations, IEEE Access 3 (2015) 2163–2175.

[26] S. Han, J. Pool, J. Tran, W. J. Dally, Learning both weights and connections for efficient neural networks, Advances in Neural Information Processing Systems (2015) 1135–1143.

[27] H. Li, A. Kadav, I. Durdanovic, H. Samet, H. P. Graf, Pruning filters for efficient convnet, International Conference on Learning Representations (2017).

[28] Y. He, X. Zhang, J. Sun, Channel pruning for accelerating very deep neural networks, IEEE International Conference on Computer Vision (2017) 1389–1397.

[29] Z. Liu, J. Li, Z. Shen, G. Huang, S. Yan, C. Zhang, Learning efficient convolutional networks through network slimming, IEEE International Conference on Computer Vision (2017) 2736–2744.

[30] T.-J. Yang, Y.-H. Chen, V. Sze, Designing energy-efficient convolutional neural networks using energy-aware pruning, IEEE Conference on Computer Vision and Pattern Recognition (2017) 5687–5695.

[31] K. Fukushima, Neocognitron: A self-organizing neural network model for a mechanism of pattern recognition unaffected by shift in position, Biological Cybernetics 36 (4) (1980) 193-202.

[32] H. Liu, X. Hou, Moving detection research of background frame difference based on gaussian model, International Conference on Computer Science and Service System (2012) 258–261.

[33] Available: https://github.com/weiliu89/caffe/tree/ssd.

[34] T.-Y. Lin, M. Maire, S. J. Belongie, L. D. Bourdev, R. B. Girshick, J. Hays, P. Perona, D. Ramanan, P. Dollár, C. L. Zitnick, Microsoft coco: common objects in context, European conference on computer vision (2014) 740–755.

[35] M. Everingham, L. Gool, C. K. Williams, J. Winn, A. Zisserman, The pascal visual object classes (voc) challenge, International Journal of Computer Vision 88 (2) (2010) 303–338.

[36] J. Ferryman, A.-L. Ellis, Performance evaluation of crowd image analysis using the pets2009 dataset, Pattern Recognition Letters 44 (2014) 3–15.

[37] E. Maggio, E. Piccardo, C. Regazzoni, A. Cavallaro, Particle phd filter for multi-target visual tracking, IEEE International Conference on Acoustics, Speech and Signal Processing 1 (2007) I–1101.

[38] S. Oh, A. Hoogs, A. Perera, N. Cuntoor, C.-C. Chen, J. T. Lee, S. Mukherjee, J. K. Aggarwal, H. Lee, L. Davis, E. Swears, X. Wang, Q. Ji, K. Reddy, M. Shah, C. Vondrick, H. Pirsiavash, D. Ramanan, J. Yuen, A. Torralba, B. Song, A. Fong, A. Roy-Chowdhury, , M. Desai, A large-scale benchmark dataset for event recognition in surveillance video, IEEE Conference on Computer Vision and Pattern Recognition (2011) 3153–3160.