深度学习算法、硬件技术及其在未来军事上的应用

doi:10.16257/j.cnki.1681-1070.2019.1201

摘要/Abstract

摘要： 人工智能在感知和认知智能领域取得的重大进展，促进该技术向军事应用的转移。从当前深度学习算法研究重点展开论述，对其未来的技术应用、发展方向进行了系统的梳理，从延续传统架构的CPU、GPU、FPGA、ASIC和非传统架构的神经拟态芯片两个阵营出发，依次对比分析了各个计算架构的结构特点以及对人工智能硬件未来发展带来的影响，同时以雷达红外图形处理技术和声呐信号处理技术在军事上的应用为例，阐述了人工智能技术在未来军事领域内的发展方向。

关键词: 深度学习, 硬件实现, 军事应用

Abstract: Significant progress has been made by artificial intelligence (AI) in the field of perceptual and cognitive intelligence, which promotes its applications in the military field. This paper discusses the current research of deep learning algorithms and systematically analyzes its application and development direction. By comparing the traditional architectures of CPU, GPU, FPGA, ASIC with non-traditional architectures of neuromorphic chips, the structural characteristics of each computing architecture and their impact on the development of AI hardware are discussed. Besides, taking the application of radar infrared image processing technology and sonar signal processing technology in the military as examples, the development direction of AI technology in the military field is analyzed.

Key words: deep learning, hardware implementation, military application

中图分类号:

TP181

魏敬和, 林军. 深度学习算法、硬件技术及其在未来军事上的应用[J]. 电子与封装, 2019, 19(12): 1 -6.

WEI Jinghe, LIN Jun. Deep Learning Algorithms, Hardware and Their Military Applications[J]. Electronics & Packaging, 2019, 19(12): 1 -6.

参考文献 40

[1]	SZEGEDY C, LIU W, JIA Y, et al. Going deeper with convolutions[C]. 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2015: 1-9.[2]HINTON G E, DENG L, YU D, et al. Deep neural networks for acoustic modeling in speech recognition: the shared views of four research groups[J]. IEEE Signal Processing　Magazine, 2012, 29(6): 82-97.
[3]	ZHU J, HUANG T, CHEN W, et al. The future of artificial intelligence in China[J]. Communications of the ACM, 2018, 61(11), 44-45.
[4]	王茜. 巨头抢滩计算机视觉入口，中国该做什么？[N]通信产业报,2017-12-11(009).
[5]	ISLAM N, ISLAM Z, NOOR N. A survey on optical character　recognition system[J]. Journal of Information & Communication Technology, 2016, 10(2): 1-4.
[6]	RIZVI M, RAZA H, TAHZEEB S, et al. Optical character recognition based intelligent database management system for examination process control[C]. 2019 16th International Bhurban Conference on Applied Sciences and Technology (IBCAST - 2019), 2019.
[7]	DENG J, GUO J, ZAFEIRIOU S, et al. ArcFace: additive angular margin loss for deep face recognition[OL].（2019-　12-09）[2019-11-26].https://arxiv.org/abs/1801.07698.
[8]	SCHERHAG U, RATHGEB C, MERKLE J, et al. Face recognition systems under morphing attacks: a survey[J]. IEEE Access, 2019: 23012-23026.
[9]	HE K, ZHANG X, REN S, et al. Deep residual learning for image recognition[C]. Computer Vision and Pattern Recognition, 2016: 770-778.
[10]	LECUN Y, BOTTOU L, BENGIO Y, et al. Gradient-based learning applied to document recognition[J]. Proceedings of the IEEE, 1998, 86(11): 2278-2324.
[11]	KRIZHEVSKY A , SUTSKEVER I , HINTON G. Imagenet　classification with deep convolutional neural networks[C]. In Advances in Neural Information Processing Systems, 2012:1097-1105.
[12]	SZEGEDY C, LIU W, JIA Y Q, et al. Going deeper with convolutions[OL]. (2014-09-17)[2019-11-26]. https://arxiv. org/abs/1409.4842.
[13]	SIMONYAN K, ZISSERMAN A. Very deep convolutional networks for large-scale image recognition[C]. International Conference on Learning Representations, 2015.
[14]	HOWARD A, ZHU M, CHEN B, et al. MobileNets: efficient　convolutional neural networks for mobile vision applications[OL]. (2017-04-17)[2019-11-26]. https://arxiv.org/abs/1704. 04861.
[15]	IANDOLA F N, HAN S, MOSKEWICZ M W, et al.SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5 MB model size[OL].(2016-11-04) [2019-11-26]. https://arxiv.org/abs/1602.07360.
[16]	COURBARIAUX M, HUBARA I, SOUDRY D, et al. Binarized neural networks: training deep neural networks with weights and activations constrained to +1 or -1[OL]. (2016-03-17) [2019-11-26]. https://arxiv.org/abs/1602.02830.
[17]	LIU W, ANGUELOV D, ERHAN D, et al. SSD: single shot MultiBox detector[C]. European Conference on Computer Vision, 2016: 21-37.
[18]	邢浩强，杜志岐，苏波. 基于改进SSD的行人检测方法[J]. 计算机工程, 2018, 44(11): 228-233, 238.
[19]	GIRSHICK R B. Fast R-CNN[C]. International Conference on Computer Vision, 2015: 1440-1448.
[20]	REN S, HE K, GIRSHICK R B, et al. Faster R-CNN: towards real-time object detection with region proposal networks[C]. Neural Information Processing Systems, 2015: 91-99.
[21]	HE K, GKIOXARI G, DOLLAR P, et al. Mask R-CNN[C]. International Conference on Computer Vision, 2017: 2980-2988.
[22]	CHEN L, ZHU Y, PAPANDREOU G, et al. Encoder-decoder　with atrous separable convolution for semantic image segmentation[C]. European Conference on Computer Vision, 2018: 833-851.
[23]	SHI B, BAI X, YAO C, et al. An end-to-end trainable neural　network for image-based sequence recognition and its application　to scene text recognition[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(11): 2298-2304.
[24]	YANG F, CHOI W, LIN Y, et al. Exploit all the layers: fast and accurate CNN object detector with scale dependent　pooling and cascaded rejection classifiers[C]. Computer Vision and Pattern Recognition, 2016: 2129-2137.
[25]	GOODFELLOW I, POUGET-ABADIE J, MIRZA M, et al. Generative adversarial nets[C]. International Conference on Neural Information Processing Systems, 2014.
[26]	RADFORD A , METZ L , CHINTALA S . Unsupervised　representation learning with deep convolutional generative International Conference on Computer Vision, 2017: 2242-　2251.
[28]	SUN K, XIAO B, LIU D, et al. Deep high-resolution representation learning for human pose estimation[C]. Computer Vision and Pattern Recognition, 2019: 5693-5703.
[29]	YANG Z P, PAN J Z, LUO L J, et al. Extreme relative pose estimation for RGB-D scans via scene completion[OL]. (2019-01-05)[2019-11-26]. https://arxiv.org/abs/1901.00063.
[30]	高晨兰，朱嘉钢. 静止背景下的人体行为识别方法[J]. 计算机工程，2017，43(10): 192-197.
[31]	CARREIRA J , ZISSERMAN A. Quo vadis, action recognition? A new model and the kinetics dataset[OL]. (2018-02-12) [2019-11-26]. https://arxiv.org/abs/1705.07750v1.
[32]	RUSU R B, COUSINS S. 3D is here: point cloud library (PCL)[C]. International Conference on Robotics and Automation, 2011: 1-4.
[33]	LORENSEN W E, CLINE, H. E. Marching cubes: a high　resolution 3D surface construction algorithm[J]. ACM SIGGRAPH Computer Graphics 21(4):163-169.
[34]	AMODEI D, HERNANDEZ D. Open AI, AI and compute[OL]. (2018-5-16)[2019-11-26]. https://openai.com/blog/ai-and-　compute.
[35]	SILVER D, HUANG A, MADDISON C J, et al. Mastering the game of go with deep neural networks and tree search[J]. Nature, 2016, 529(7587): 484-489.
[36]	JOUPPI N P, YOUNG C S, Patil N, et al. In-datacenter performance analysis of a tensor processing unit[J]. International Symposium on Computer Architecture, 2017, 45(2): 1-12.
[37]	MEROLLA P A, ARTHUR J V, ALVAREZICAZA R, et al. A million spiking-neuron integrated circuit with a scalable communication network and interface[J]. Science, 2014, 345(6197): 668-673.
[38]	孙欣. 现代智能作战概念浅析[J]. 舰船电子工程，2009, 29(8):9-13.
[39]	吴世龙，吴世岭. 电子战武器装备的发展趋势与发展重点[J]. 舰船电子对抗, 2004, 27(6):3-5.
[40]	董磊.世界进入人工智能战争时代：机器自主杀人[OL]. (2017-06-02)[2019-11-26]. http://www.sohu.com/a/115401680_465915. International Conference on Computer Vision, 2017: 2242-　2251.

中国半导体行业协会封装分会会刊

中国电子学会电子制造与封装技术分会会刊