A multi-modal multi-scale network based on Transformer for micro-expression recognition

doi:10.1016/j.jvcir.2025.104537

A multi-modal multi-scale network based on Transformer for micro-expression recognition

2025-07-11

0

PRE

AI

DOI：10.1016/j.jvcir.2025.104537

Original

Original request for help

Save

Abstract

En 中文

• Two complementary micro-expression features extracted from different modalities. • Multi-scale attention helps capture important local features at different scales. • Cross-entropy and unsupervised contrastive learning loss to constrain the model.

A multi-modal multi-scale network based on Transformer for micro-expression recognition

Abstract

Journal

Organization