A modified contrastive loss method for face recognition

Abstract Contrastive Loss is frequently used as loss function in CNN for face recognition, but it can result in the overfitting and low sampling efficiency for the positive samples. In this paper, a Modified Contrastive Loss (MCL) is proposed to overcome the shortcomings of contrastive loss. MCL and ResNet are combined with a Joint Bayesian technique to develop a ResNet-Modified Contrastive Loss-Joint Bayesian (ResNet-MCL-JB) model. First, ResNet is used as the basic network structure, and several ResNets are trained to use the MCL. Then, the ResNet with the Joint Bayesian for metric learning is integrated. The state-of-the-art performance of ResNet-MCL-JB attests to its effect. For further improvement, a Progressive Soft Filter Pruning method (PSFP) is applied in the neural network. PSFP can effectively diminish the size of the network while maintaining high accuracy. This method gradually prunes the filters on each layer by the weight of each filter. We combine the MCL and PSFP together with ResNet, and thus, we achieve considerable much improvement in both accuracy and computational cost.Keywords: Modified contrastive loss; Face recognition; Joint bayesian

[1]  Muriel Visani,et al.  Simple Triplet Loss Based on Intra/Inter-Class Metric Learning for Face Verification , 2017, 2017 IEEE International Conference on Computer Vision Workshops (ICCVW).

[2]  Kihyuk Sohn,et al.  Improved Deep Metric Learning with Multi-class N-pair Loss Objective , 2016, NIPS.

[3]  James Philbin,et al.  FaceNet: A unified embedding for face recognition and clustering , 2015, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[4]  Jiwen Lu,et al.  Sharable and Individual Multi-View Metric Learning , 2018, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[5]  Yann LeCun,et al.  Learning a similarity metric discriminatively, with application to face verification , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).

[6]  Jiwen Lu,et al.  Deep Metric Learning for Visual Understanding: An Overview of Recent Advances , 2017, IEEE Signal Processing Magazine.

[7]  Xudong Lin,et al.  Deep Adversarial Metric Learning , 2018, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[8]  Andrew Zisserman,et al.  Deep Face Recognition , 2015, BMVC.

[9]  Yu Qiao,et al.  A Discriminative Feature Learning Approach for Deep Face Recognition , 2016, ECCV.

[10]  Jiwen Lu,et al.  Discriminative Deep Metric Learning for Face and Kinship Verification , 2017, IEEE Transactions on Image Processing.

[11]  Fei Su,et al.  Contrastive-center loss for deep neural networks , 2017, 2017 IEEE International Conference on Image Processing (ICIP).

[12]  Jian Sun,et al.  Bayesian Face Revisited: A Joint Formulation , 2012, ECCV.

[13]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[14]  Jiwen Lu,et al.  Deep Localized Metric Learning , 2018, IEEE Transactions on Circuits and Systems for Video Technology.