Angular Sparsemax for Face Recognition

The Softmax prediction function is widely used to train Deep Convolutional Neural Networks (DCNNs) for large-scale face recognition and other applications. The limitation of the softmax activation is that the resulting probability distribution always has a full support. This full support leads to larger intraclass variations. In this paper, we formulate a novel loss function, called Angular Sparsemax for face recognition. The proposed loss function promotes sparseness of the hypotheses prediction function similar to Sparsemax [1] with Fenchel-Young regularisation. By introducing an additive angular margin on the score vector, the discriminatory power of the face embedding is further improved. The proposed loss function is experimentally validated on several databases in terms of recognition accuracy. Its performance compares well with the state of the art Arcface loss.

[1]  Stefanos Zafeiriou,et al.  AgeDB: The First Manually Collected, In-the-Wild Age Database , 2017, 2017 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW).

[2]  C. Tsallis.  Possible generalization of Boltzmann-Gibbs statistics , 1988 .

[3]  Feiyue Huang,et al.  CurricularFace: Adaptive Curriculum Learning Loss for Deep Face Recognition , 2020, 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[4]  Kaiming He,et al.  Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour , 2017, ArXiv.

[5]  Shengcai Liao,et al.  Learning Face Representation from Scratch , 2014, ArXiv.

[6]  Lanqing He,et al.  Softmax Dissection: Towards Understanding Intra- and Inter-clas Objective for Embedding Learning , 2020, AAAI.

[7]  Carlos D. Castillo,et al.  Crystal Loss and Quality Pooling for Unconstrained Face Verification and Recognition , 2018, ArXiv.

[8]  Bhiksha Raj,et al.  SphereFace: Deep Hypersphere Embedding for Face Recognition , 2017, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[9]  Carlos D. Castillo,et al.  Frontal to profile face verification in the wild , 2016, 2016 IEEE Winter Conference on Applications of Computer Vision (WACV).

[10]  Stefanos Zafeiriou,et al.  RetinaFace: Single-stage Dense Face Localisation in the Wild , 2019, ArXiv.

[11]  Zhi Zhang,et al.  Bag of Tricks for Image Classification with Convolutional Neural Networks , 2018, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[12]  Xing Ji,et al.  CosFace: Large Margin Cosine Loss for Deep Face Recognition , 2018, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[13]  Shifeng Zhang,et al.  Mis-classified Vector Guided Softmax Loss for Face Recognition , 2019, AAAI.

[14]  Ira Kemelmacher-Shlizerman,et al.  The MegaFace Benchmark: 1 Million Faces for Recognition at Scale , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[15]  André F. T. Martins,et al.  Learning Classifiers with Fenchel-Young Losses: Generalized Entropies, Margins, and Algorithms , 2018, AISTATS.

[16]  Yang Liu,et al.  MobileFaceNets: Efficient CNNs for Accurate Real-time Face Verification on Mobile Devices , 2018, CCBR.

[17]  Ramón Fernández Astudillo,et al.  From Softmax to Sparsemax: A Sparse Model of Attention and Multi-Label Classification , 2016, ICML.

[18]  Yuxiao Hu,et al.  MS-Celeb-1M: A Dataset and Benchmark for Large-Scale Face Recognition , 2016, ECCV.

[19]  Debing Zhang,et al.  Lightweight Face Recognition Challenge , 2019, 2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW).

[20]  Xiaogang Wang,et al.  Deep Learning Face Representation from Predicting 10,000 Classes , 2014, 2014 IEEE Conference on Computer Vision and Pattern Recognition.

[21]  Stefanos Zafeiriou,et al.  ArcFace: Additive Angular Margin Loss for Deep Face Recognition , 2018, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[22]  Marwan Mattar,et al.  Labeled Faces in the Wild: A Database forStudying Face Recognition in Unconstrained Environments , 2008 .

[23]  Naonori Ueda,et al.  Large-Scale Multiclass Support Vector Machine Training via Euclidean Projection onto the Simplex , 2014, 2014 22nd International Conference on Pattern Recognition.

[24]  Jian Cheng,et al.  Additive Margin Softmax for Face Verification , 2018, IEEE Signal Processing Letters.

[25]  Miguel Á. Carreira-Perpiñán,et al.  Projection onto the probability simplex: An efficient algorithm with a simple proof, and an application , 2013, ArXiv.