July 5, 2023
Using Deep Learning for Speech Emotion Recognition
Machine learning
To enhance Speech Emotion Recognition (SER) through Deep Convolutional Neural Networks (DCNNs), we used the fast Continuous Wavelet Transform (fCWT). Its time-frequency analysis offers real-time CWT, providing improved results over the commonly used Short-Term Fourier Transform (STFT) based SER systems. We show that fCWT-DCNN-based SER shows promise for real-time high-accuracy SER for a wide range of applications.