Emotion Recognition with Facial Attention and Objective Activation Functions

TL;DR

It is shown that not only attention can significantly improve the performance of these models but also that combining them with a different activation function can further help increase the performance of these models.

Abstract

In this paper, we study the effect of introducing channel and spatial attention mechanisms, namely SEN-Net, ECA-Net, and CBAM, to existing CNN vision-based models such as VGGNet, ResNet, and ResNetV2 to perform the Facial Emotion Recognition task. We show that not only attention can significantly improve the performance of these models but also that combining them with a different activation function can further help increase the performance of these models.

Figures (2)

• Figure 1: raining and validation accuracy graphs of ResNet(left) and ResNetV2(right) with 3 Different Depths (50, 101 and 152) on the JAFFE dataset.
• Figure 2: Accuracy curves for the best performing models on the CK+(left), JAFFE(middle), and FER2013(right).