FaceMixup: Enhancing Facial Expression Recognition through Mixed Face Regularization

August 2026

The proliferation of deep learning solutions and the scarcity of large annotated datasets pose significant challenges in real-world appli- cations. Various strategies have been explored to overcome this challenge, with data augmentation (DA) approaches emerging as prominent solu- tions. DA approaches involve generating additional examples by trans- forming existing labeled data, thereby enriching the dataset and helping deep learning models achieve improved generalization without succumb- ing to overfitting. In real applications, where solutions based on deep learning are widely used, there is facial expression recognition (FER), which plays an essential role in human communication, improving a range of knowledge areas (e.g., medicine, security, and marketing). In this pa- per, we propose a simple and comprehensive face data augmentation ap- proach based on mixed face component regularization that outperforms the classical DA approaches from the literature, including the MixAug- ment which is a specific approach for the target task in two well-known FER datasets existing in the literature.

Authors in the community:

• 

  F

  Fabio Augusto Faria

  ist430285

International Conference on Pattern Recognition 2026

computer-and-information-sciences - Computer and information sciences