DCMA-CNNs

a demo code of model DCMA-CNNs

Brief Introduction

This is the code with respect to the model DCMA-CNNs, which is proposed in the paper: <Facial Expression Recognition Using Hierarchical Features With Deep Comprehensive Multipatches Aggregation Convolutional Neural Networks>

Settings

• The model is developed in MATLAB. Experiments are conducted on MATLAB 2014a.
• Experiments were all run on hardware with 8GB RAM and a single 2.5GHz CPU.
• The code is based on the DeepLearnToolBox

Data

• Experiments (details refer to the paper) are conducted on two facial expression datasets: CK+ and JAFFE. You can apply for or directly download these two dataset.
• This code is just a demo of DCMA-CNNs. I only include six samples (all sampled from the JAFFE dataset) to show how DCMA-CNNs works (training and evaluation).