Research on Facial Expression Recognition Based on VGG16 Network Model
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
Facial expression recognition (FER), as an interdisciplinary field of emotion computing and computer vision, aims to automatically recognize the emotional states contained in human facial expressions through computers. This article proposes a research on facial expression recognition based on an improved VGG16 network model to address the problem of traditional algorithms relying on manual design for feature extraction and being unable to extract deep features. The network performance is optimized by introducing channel attention mechanism and Batch Normalization (BN). The experimental results show that the improved model achieved an accuracy of 64.70% on the FER2013 test set. In addition, this article analyzed the recognition performance of the model on various expression categories through confusion matrix and explored possible optimization directions in the future.
Fields of Study & Subjects
Emotion and Mood Recognition
Face recognition and analysis
Face and Expression Recognition
Publication details
- Journal / source
- IEEE International Conference on Power, Electronics and Green Energy (ICPEGE)
- Pages
- 1559–1562
- Publication date
- 28 January 2026
- Language
- EN