# STRCN-MicroExpressionRec

**Repository Path**: xia_zhaoqiang/STRCN-MicroExpressionRec

## Basic Information

- **Project Name**: STRCN-MicroExpressionRec
- **Description**: This repository contains the source code for our TMM work, which is entitled as "Spatiotemporal Recurrent Convolutional Networks for Recognizing Spontaneous Micro-Expressions". 
- **Primary Language**: Matlab
- **License**: CC-BY-4.0
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2021-12-31
- **Last Updated**: 2021-12-31

## Categories & Tags

**Categories**: Uncategorized

**Tags**: MER

## README

# STRCN-MicroExpressionRec

Spatiotemporal Recurrent Convolutional Networks for Recognizing Spontaneous Micro-Expressions
Descriptions

These codes are used for micro-expression recognition on individual datasets. The methods can be accessed by the paper "Spatiotemporal Recurrent Convolutional Networks for Recognizing Spontaneous Micro-Expressions, IEEE TMM2020", which includes the STRCN-A and STRCN-G.
Dependencies

The code was written in MATLAB R2018b, and tested on Windows 7 and 10.

    MATLAB's Computer Vision Toolbox
    MatConvNet (Beta23)
    Eulerian Video Magnification (2012):http://people.csail.mit.edu/mrub/vidmag/

Instructions

    The faces should be cropped by your own tools and applified by the EVM.
    All images are storaged with the format of casme2 way.
    The appearance connect and geometric connect needs to be performed before training the deep model. All the data needs to be prepared and storaged into the mat file.