US 12,354,405 B1
Expression recognition method and system based on multi-scale features and spatial attention
Zhaowei Liu, Yantai (CN); Haonan Wen, Yantai (CN); Yongchao Song, Yantai (CN); Wenhan Hou, Yantai (CN); Xinxin Zhao, Yantai (CN); Tengjiang Wang, Yantai (CN); Diantong Liu, Yantai (CN); Weiqing Yan, Yantai (CN); Peng Song, Yantai (CN); Anzuo Jiang, Yantai (CN); and Hang Su, Yantai (CN)
Assigned to YANTAI UNIVERSITY, Yantai (CN)
Filed by YANTAI UNIVERSITY, Yantai (CN)
Filed on Feb. 24, 2025, as Appl. No. 19/061,635.
Application 19/061,635 is a continuation of application No. PCT/CN2024/135203, filed on Nov. 28, 2024.
Claims priority of application No. 202410710860.8 (CN), filed on Jun. 4, 2024.
Int. Cl. G06V 40/16 (2022.01); G06V 10/82 (2022.01)
CPC G06V 40/175 (2022.01) [G06V 10/82 (2022.01); G06V 40/171 (2022.01)] 10 Claims
OG exemplary drawing
 
1. An expression recognition method based on multi-scale features and spatial attention, comprising:
acquiring facial image data;
constructing an HNFER neural network model;
performing feature extraction on acquired facial image data by using the HNFER neural network model to obtain an original input feature map;
performing pooling and concatenation on extracted features based on a CoordAtt attention mechanism to obtain a feature map;
performing deep convolution processing on the feature map to obtain an attention map, and then performing element-by-element multiplication to obtain a final feature map; and
performing feature transformation and normalization on the final feature map to obtain an expression category probability and output the expression category probability.