Face Expression Recognition Method Based on Mamba and Different Facial Regions

With the rapid advancement of information technology, facial expression recognition has become increasingly critical in various applications. To enhance recognition accuracy and computational efficiency, we propose a novel integration of the Vision Mamba framework with facial part-based image classification. This approach precisely segments key facial regions using Grad-CAM heatmaps, enabling the Mamba model to extract and analyze deep features from critical areas. By leveraging these techniques, our method achieves superior classification performance across multiple datasets. Experimental results demonstrate its robustness and efficiency, making it highly suitable for facial expression recognition. This work provides a practical and scalable solution, paving the way for advancements in intelligent systems and broader application scenarios, such as human-computer interaction and affective computing.