Learning Deep Features via Congenerous Cosine Loss for Person Recognition

Abstract

Person recognition aims at recognizing the same identity across time and space with complicated scenes and similar appearance. In this paper, we propose a novel method to address this task by training a network to obtain robust and representative features. The intuition is that we directly compare and optimize the cosine distance between two features - enlarging inter-class distinction as well as alleviating inner-class variance. We propose a congenerous cosine loss by minimizing the cosine distance between samples and their cluster centroid in a cooperative way. Such a design reduces the complexity and could be implemented via softmax with normalized inputs. Our method also differs from previous work in person recognition that we do not conduct a second training on the test subset. The identity of a person is determined by measuring the similarity from several body regions in the reference set. Experimental results show that the proposed approach achieves better classification accuracy against previous state-of-the-arts.

arXiv commentPost-rebuttal update. Add some comparison results; correct some technical part; rewrite some sections to make it more readable; code link available

References in corpus (3)

• Unsupervised Learning of Visual Representations using Videos
• Dual Deep Network for Visual Tracking
• Faceless Person Recognition; Privacy Implications in Social Media

Cited by in corpus (20)

• Recent Advances in Deep Learning Techniques for Face Recognition
• Support Vector Guided Softmax Loss for Face Recognition
• AdaCos: Adaptively Scaling Cosine Logits for Effectively Learning Deep Face Representations
• SphereFace2: Binary Classification is All You Need for Deep Face Recognition
• A Performance Comparison of Loss Functions for Deep Face Recognition
• Regular Polytope Networks
• Adversarial Image Perturbation for Privacy Protection -- A Game Theory Perspective
• Feature Incay for Representation Regularization
• Ring loss: Convex Feature Normalization for Face Recognition
• Fix Your Features: Stationary and Maximally Discriminative Embeddings using Regular Polytope (Fixed Classifier) Networks
• Towards Interpretable Face Recognition
• Clustering-Oriented Representation Learning with Attractive-Repulsive Loss
• Zoom Out-and-In Network with Map Attention Decision for Region Proposal and Object Detection
• Recurrent Scale Approximation for Object Detection in CNN
• Ensemble Soft-Margin Softmax Loss for Image Classification
• Worse WER, but Better BLEU? Leveraging Word Embedding as Intermediate in Multitask End-to-End Speech Translation
• Query Adaptive Late Fusion for Image Retrieval
• Deep Class-Wise Hashing: Semantics-Preserving Hashing via Class-wise Loss
• Anchor-based Nearest Class Mean Loss for Convolutional Neural Networks
• Exploiting Web Images for Fine-Grained Visual Recognition by Eliminating Noisy Samples and Utilizing Hard Ones