Learning label correlations for multi-label image recognition with graph networks

Qing Li, Xiaojiang Peng, Yu Qiao, Qiang Peng

Abstract: Highlights•We propose an A-GCN framework for multi-label image recognition.•We introduce a plug-and-play adaptive label graph (LG) module to automatically learn the label correlations.•We explore three different LG architectures for effective evaluated the LG module.•We propose a spares correlation constraint to enhance the LG module.