Abstract | ||
---|---|---|
•The novel end-to-end architecture can improve the accuracy of action recognition efficiently.•Introducing the part alignment into action recognition can capture spatio-temporal evolutions of actions.•The part-based hierarchical pooling approach can learn a robust and discriminative feature.•Our method obtains the state-of-the-art results on two important benchmarks of action recognition. |
Year | DOI | Venue |
---|---|---|
2019 | 10.1016/j.patcog.2019.03.010 | Pattern Recognition |
Keywords | Field | DocType |
Action recognition,Part alignment,Auto-transformer attention | Complementarity (molecular biology),Body Representation,Pattern recognition,Spatial configuration,Action recognition,Pooling,Exploit,Artificial intelligence,Machine learning,Feature learning,Human body,Mathematics | Journal |
Volume | Issue | ISSN |
92 | 1 | 0031-3203 |
Citations | PageRank | References |
1 | 0.34 | 0 |
Authors | ||
4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Linjiang Huang | 1 | 6 | 3.14 |
Yan Huang | 2 | 226 | 27.65 |
Wanli Ouyang | 3 | 2371 | 105.17 |
Liang Wang | 4 | 4317 | 243.28 |