EUVIMEDEUVIMED entdecken
Crossref

PASTFNet: a paralleled attention spatio-temporal fusion network for micro-expression recognition

Haichen Tian, Weijun Gong, Wei Li, Yurong Qian

Medical & Biological Engineering & Computing · 2024

DOI: 10.1007/s11517-024-03041-y ↗

Kurz gefasst

Fachartikel in der Zeitschrift Medical & Biological Engineering & Computing, 2024. Verfasst von Haichen Tian und 3 weiteren Autor:innen. Beteiligte Einrichtung: School of Information Science and Engineering, Xinjiang University, Urumqi, China. Hauptthemen laut MEDLINE: Neural Networks, Computer. Die Arbeit wurde bisher 11-mal zitiert (Crossref) und nennt 44 Literaturangaben. Gefördert von: The National Science Foundation of China, National Science and Technology Major Project, Autonomous Region Science and Technology International Cooperation Project.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Micro-expressions (MEs) play such an important role in predicting a person's genuine emotions, as to make micro-expression recognition such an important resea rch focus in recent years. Most recent researchers have made efforts to recognize MEs with spatial and temporal information of video clips. However, because of their short duration and subtle intensity, capturing spatio-temporal features of micro-expressions remains challenging. To effectively promote the recognition performance, this paper presents a novel paralleled dual-branch attention-based spatio-temporal fusion network (PASTFNet). We jointly extract short- and long-range spatial relationships in spatial branch. Inspired by the composite architecture of the convolutional neural network (CNN) and long short-term memory (LSTM) for temporal modeling, we propose a novel attention-based multi-scale feature fusion network (AMFNet) to encode features of sequential frames, which can learn more expressive facial-detailed features for it implements the integrated use of attention and multi-scale feature fusion, then design an aggregation block to aggregate and acquire temporal features. At last, the features learned by the above two branches are fused to accomplish expression recognition with outstanding effect. Experiments on two MER datasets (CASMEII and SAMM) show that the PASTFNet model achieves promising ME recognition performance compared with other methods.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Neural Networks, ComputerHumansAttentionFacial ExpressionEmotionsAlgorithmsImage Processing, Computer-AssistedPattern Recognition, Automated

Stichwörter der Autor:innen

AttentionDual-branchMicro-expression recognitionMulti-scale fusionSpatio-temporal features

Autor:innen und Institutionen

  • School of Information Science and Engineering, Xinjiang University, Urumqi, China.
  • School of Software, Xinjiang University, Urumqi, China.
  • School of Information Science and Engineering, Xinjiang University, Urumqi, China. qyr@xju.edu.cn.
  • School of Software, Xinjiang University, Urumqi, China. qyr@xju.edu.cn.
  • Key Laboratory of Signal Detection and Processing in Xinjiang Uygur Autonomous Region, Urumqi, China. qyr@xju.edu.cn.

Förderung

  • The National Science Foundation of China
  • National Science and Technology Major Project
  • Autonomous Region Science and Technology International Cooperation Project

Zitieren

APA

Tian, H., Gong, W., Li, W., & Qian, Y. (2024). PASTFNet: a paralleled attention spatio-temporal fusion network for micro-expression recognition. Medical & Biological Engineering & Computing. https://doi.org/10.1007/s11517-024-03041-y

Vancouver

Tian H, Gong W, Li W, Qian Y. PASTFNet: a paralleled attention spatio-temporal fusion network for micro-expression recognition. Medical & Biological Engineering & Computing. 2024. doi:10.1007/s11517-024-03041-y

BibTeX
@article{Tian2024,
  author = {Tian, Haichen and Gong, Weijun and Li, Wei and Qian, Yurong},
  title = {PASTFNet: a paralleled attention spatio-temporal fusion network for micro-expression recognition},
  journal = {Medical & Biological Engineering & Computing},
  year = {2024},
  doi = {10.1007/s11517-024-03041-y}
}
Kostenlos starten →