Sensitive deep convolutional neural network for face recognition at large standoffs with small dataset

作者：

Highlights：

• Incorporation of sensitivity term in cost function of deep CNN structure.

• Slight variations and high frequency components are emphasized by sensitivity term.

• Sensitivity pushes samples at boundaries to high gradient part of activation.

• Application to unconstrained environments with small sample size.

摘要

•Incorporation of sensitivity term in cost function of deep CNN structure.•Slight variations and high frequency components are emphasized by sensitivity term.•Sensitivity pushes samples at boundaries to high gradient part of activation.•Application to unconstrained environments with small sample size.

论文关键词：Convolutional neural network,Gradient descent,Input-output mapping sensitivity error back propagation,Face recognition at long distances with small dataset,Sensitivity in cost function,Deep neural structures

论文评审过程：Received 1 December 2016, Revised 14 June 2017, Accepted 15 June 2017, Available online 16 June 2017, Version of Record 23 June 2017.

论文官网地址：https://doi.org/10.1016/j.eswa.2017.06.025