Abstract:In demanding application scenarios such as clinical psychotherapy and criminal interrogation, the accurate recognition of micro-expressions is of utmost importance but poses significant challenges. One of the main difficulties lies in effectively capturing weak and fleeting facial features and improving recognition performance. To address this fundamental issue, this paper proposed a novel architecture based on a multi-scale 3D residual convolutional neural network. The algorithm leveraged a deep 3D-ResNet50 as the skeleton model and utilized the micro-expression optical flow feature map as the input for the network model. Drawing upon the complex spatial and temporal features inherent in micro-expressions, the network incorporated multi-scale convolutional modules of varying sizes to integrate both global and local information. Furthermore, an attention mechanism feature fusion module was introduced to enhance the model's contextual awareness. Finally, to optimize the model's prediction of the optimal solution, a discriminative network structure with multiple output channels was constructed. The algorithm's performance was evaluated using the public datasets SMIC, SAMM, and CASME II. The experimental results demonstrated that the proposed algorithm achieves recognition accuracies of 74.6, 84.77 and 91.35% on these datasets, respectively. This substantial improvement in efficiency compared to existing mainstream methods for extracting micro-expression subtle features effectively enhanced micro-expression recognition performance and increased the accuracy of high-precision micro-expression recognition. Consequently, this paper served as an important reference for researchers working on high-precision micro-expression recognition.

Selective Deep Features for Micro-Expression Recognition

Micro-Attention for Micro-Expression recognition

Micro-Facial Expression Recognition Based on Deep-Rooted Learning Algorithm

Spatiotemporal Recurrent Convolutional Networks for Recognizing Spontaneous Micro-expressions

Facial Micro-Expression Recognition Based on Deep Local-Holistic Network

Facial Micro Expression Detection Using Deep Learning Architecture

3D Feature Extraction Network Based on Self-supervision for Micro-expression Spotting

GEME: Dual-stream Multi-Task GEnder-based Micro-Expression Recognition.

A Survey of Micro-expression Recognition Methods Based on LBP, Optical Flow and Deep Learning

Spontaneous Facial Micro-expression Recognition Via Deep Convolutional Network

MCCA-VNet: A Vit-Based Deep Learning Approach for Micro-Expression Recognition Based on Facial Coding

Block Division Convolutional Network with Implicit Deep Features Augmentation for Micro-Expression Recognition

Micro-expression recognition based on multi-scale 3D residual convolutional neural network

Deep Facial Expression Recognition Using Transfer Learning and Fine-Tuning Techniques

Dual-Stream Shallow Networks For Facial Micro-Expression Recognition

Deep Learning for Micro-Expression Recognition: A Survey

Enriched Long-Term Recurrent Convolutional Network for Facial Micro-Expression Recognition

Analysis of Micro Facial Expression by Machine and Deep Learning Methods: Haar, CNN, and RNN

The heterogeneous ensemble of deep forest and deep neural networks for micro-expressions recognition

Deep Evolution for Facial Emotion Recognition

Micro-Expression Recognition Based on Optical Flow and PCANet+