[1] [1] NVIDIA. NVIDIA Deep learning SDK[DBOL]. [20191127] https:docs.nvidia.comdeeplearningsdkindex.html.

[2] G E HINTON. Where do features come from?. Cognitive Science, 38, 1078-1101(2014).

[3] [3] BISHOP C.Neutal wks f pattern recognition[M]. London: Oxfd University Press, 1995.

[4] [4] LECUN Y, BOTTOU L, BENGIO Y, et al. Gradientbased learning applied to document recognition[J]. Proceedings of the IEEE, 1998, 86:22782324.

[5] [5] RID S, CHRISTOPHER M, REW N. Learning continuous phrase representations syntactic parsing with recursive neural wks[C]Taboe Nevada: MPS Neural Infmation Processing System, 2010: 19.

[6] Y CHEN, T CHEN, UZ X. Diannao family: energy-efficient hardware accelerators for machine learning. Communications of the ACM, 59, 105-112(2016).

[7] [7] SUN F, WANG C, GONG L, et al. A powerefficient accelerat f convolutional neural wks[C]2017 IEEE International Conference on Cluster Computering (CLUSTER). Washington: IEEE, 2017: 631632.

[8] T LUO, S LIU, L LI. Dadiannao: a neural network supercomputer. IEEE Transactions on Computers, 66, 73-88(2017).

[9] [9] GOKEN E, ERASLAN Z A, JULIEN G, et al. Deep learning: new computational modelling techniques f genomics[J]. Nature Reviews Geics, 2019, 20(7): 389403.

[10] [10] AGARWAL A, DUCHI J C. Distributed delayed stochastic optimization[C]Nevada: MIPS, Neural Infmation Processing Systems, 2011: 873881.

[11] [11] YOU Y, ZHANG Z, HSIEH CJ, et al. Image training in minutes [C]Proceedings of the 47th International Conference on Parallel Processing. New Yk: NY ACM, 2018: 110.

[12] D SILVER, J SCHRITTWIESER, K SIMONYAN. Mastering the game of go without human knowledge. Nature, 550, 354(2017).

[13] [13] LONG J, SHELHAMER E, DARRELL T. Fully convolutional wks f semantic segmentation[C]Proceeedings of 2015 IEEE Conference on Computer Vision Pattern Recongnition. Boston, Washington: IEEE, 2015: 34113440.

[14] , 16, 4694-4702(2015).

[15] [15] YANG Z, NEVATIA R. A multiscale cade fully convolutional wk face detect[C]Pattern Recognition(ICPR), 2016 23rd International Conference on IEEE. Washington: IEEE, 2016: 633638.