Overview:

Facial expression recognition is a challenging task, especially when dealing with spontaneous micro-expressions. This paper introduces two 3D-CNN methods, MicroExpSTCNN and MicroExpFuseNet, for recognizing spontaneous facial micro-expressions by leveraging spatiotemporal information in CNN framework. The proposed MicroExpSTCNN model demonstrates superior performance compared to state-of-the-art methods, showcasing advancements in this field.

Features:

Spatiotemporal Information: Utilizes spatiotemporal information to recognize spontaneous facial micro-expressions.
MicroExpSTCNN: Considers full spatial information for accurate recognition.
MicroExpFuseNet: Utilizes 3D-CNN feature fusion of eyes and mouth regions to enhance recognition accuracy.
Performance Improvement: Outperforms existing methods in recognizing facial micro-expressions.

Summary:

This paper presents the development of two 3D-CNN methods, MicroExpSTCNN and MicroExpFuseNet, for the recognition of spontaneous facial micro-expressions. By leveraging spatiotemporal information and feature fusion techniques, the proposed models showcase improved performance over existing methods. The advancements in this research area provide a promising step towards more accurate and reliable facial expression recognition systems for various applications.

Bogireddytejareddy

Micro Expression Recognition