A breakthrough in Speech emotion recognition using Deep Retinal Convolution Neural Networks

TL;DR

This paper introduces a novel data augmentation method inspired by retinal imaging principles and a Deep Retinal Convolution Neural Network (DRCNN) that significantly improves speech emotion recognition accuracy, surpassing previous methods.

Contribution

The paper presents a new data augmentation technique based on retinal imaging principles and a specialized deep learning model, DRCNN, for enhanced speech emotion recognition.

Findings

Achieved over 99% average accuracy in SER

Outperformed previous methods in recognizing multiple emotions

Demonstrated effectiveness of retinal-inspired data augmentation

Abstract

Speech emotion recognition (SER) is to study the formation and change of speaker's emotional state from the speech signal perspective, so as to make the interaction between human and computer more intelligent. SER is a challenging task that has encountered the problem of less training data and low prediction accuracy. Here we propose a data augmentation algorithm based on the imaging principle of the retina and convex lens, to acquire the different sizes of spectrogram and increase the amount of training data by changing the distance between the spectrogram and the convex lens. Meanwhile, with the help of deep learning to get the high-level features, we propose the Deep Retinal Convolution Neural Networks (DRCNNs) for SER and achieve the average accuracy over 99%. The experimental results indicate that DRCNNs outperforms the previous studies in terms of both the number of emotions and the accuracy of recognition. Predictably, our results will dramatically improve human-computer interaction.