Abstract | ||
---|---|---|
AbstractAbstractRecent studies use Graph Convolution Networks (GCN) to model label correlation for multi-label images because of the outstanding performance of GCN in relational modeling tasks. However, the traditional GCN has low generalization, and the current state-of-the-arts' accuracy is poor. Therefore, we propose a Two-Stream Dynamic Graph Convolution Network (2S-DGCN) to improve the performance of multi-label image recognition. In 2S-DGCN, we first obtain the Up Confidence Score of prediction categories (UCS), the content-aware category and the label discriminant vector by a Semantic Attention Module (SAM) and a Dynamic Graph Convolution Network (DGCN) in upstream. Then fed the new graph feature nodes reconstructed by lateral embedding the content-aware category and the label discriminant vector into a DGCN to produce the Down Confidence Score of prediction categories (DCS) in downstream. Finally, the Final Confidence Score of prediction categories (FCS) for multi-label image recognition is synthesized by fusing the UCS and DCS. Extensive experiments on the public multi-label benchmarks achieve mAPs of 85.6% on MS-COCO and 95.4% on VOC 2007. The results of compared experiment and visualization demonstrate that our method has better performance than the current state-of-the-art methods. |
Year | DOI | Venue |
---|---|---|
2021 | 10.1016/j.imavis.2021.104238 | Periodicals |
Keywords | DocType | Volume |
Multi-label image recognition, Two streams, Reconstructing graph feature nodes, Dynamic graph convolution networks | Journal | 113 |
Issue | ISSN | Citations |
C | 0262-8856 | 1 |
PageRank | References | Authors |
0.37 | 0 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Pingping Cao | 1 | 1 | 0.37 |
Pengpeng Chen | 2 | 123 | 17.75 |
Qiang Niu | 3 | 8 | 7.67 |