MHSA-Net: Multihead Self-Attention Network for Occluded Person Re-Identification

2020/08/31 by Hongchen Tan, Xiuping Liu, Baocai Yin +1 · 189 citations
Computer Science · #Artificial intelligence #Computer science #Computer security #Computer vision #Face recognition and analysis #Filter (signal processing) #Human Pose and Action Recognition #Identification (biology) #Image (mathematics) #Information retrieval #Key (lock) #Noise (video) #Video Surveillance and Tracking Methods #cs.CV

paper · pdf · doi:10.1109/tnnls.2022.3144163

published in IEEE Transactions on Neural Networks and Learning Systems 34(11), 8210-8224 (Institute of Electrical and Electronics Engineers) · Accepted by TNNLS

openalex publication_date 2022/03/21 · openalex created_date 2022/04/03 · arxiv created 2022/04/16 · arxiv updated 2022/04/19 · openalex updated_date 2026/08/06

Abstract

This article presents a novel person reidentification model, named multihead self-attention network (MHSA-Net), to prune unimportant information and capture key local information from person images. MHSA-Net contains two main novel components: multihead self-attention branch (MHSAB) and attention competition mechanism (ACM). The MHSAB adaptively captures key local person information and then produces effective diversity embeddings of an image for the person matching. The ACM further helps filter out attention noise and nonkey information. Through extensive ablation studies, we verified that the MHSAB and ACM both contribute to the performance improvement of the MHSA-Net. Our MHSA-Net achieves competitive performance in the standard and occluded person Re-ID tasks.

Citations

• Adam: A Method for Stochastic Optimization
• Neural Machine Translation by Jointly Learning to Align and Translate
• A Structured Self-attentive Sentence Embedding
• In Defense of the Triplet Loss for Person Re-Identification
• Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour
• Random Erasing Data Augmentation
• Deep Spatial Feature Reconstruction for Partial Person Re-identification: Alignment-Free Approach
• Beyond Part Models: Person Retrieval with Refined Part Pooling (and a Strong Convolutional Baseline)
• FD-GAN: Pose-guided Feature Distilling GAN for Robust Person Re-identification
• All You Need is Beyond a Good Init: Exploring Better Solution for Training Extremely Deep Convolutional Neural Networks with Orthonormality and Modulation
• Discriminative Feature Learning With Foreground Attention for Person Re-Identification
• On the Relationship between Self-Attention and Convolutional Layers
• Random Erasing Data Augmentation
• Deep Learning for Person Re-Identification: A Survey and Outlook
• Deep Learning for Person Re-identification: A Survey and Outlook