Speech Emotion Recognition Using LSTM Neural Network Model

Abstract

Speech Emotion Recognition SER has become a widely studied issue in the circle of affective computing. It has been used in human-computer interaction and neuropsychology. In the current research, deep learning methodologies-based models are constructed for artificial intelligence integration into SER systems. We use the neural network form of Long Short-Term Memory (LSTM) for automatic emotion detection and classification using audio data recordings. We introduce a method that preprocesses MFCCs and audio data as features and subsequently uses LSTM for classification and learning. Using the model, training as well as testing is conducted over a dataset that has numerous examples of diverse emotions. The effectiveness of our approach is manifested by the whole experiment and the test, which shows the correctness of the suggested model in the exact emotion recognition from speech. The results indicate the consequence of the outstanding performance, high accuracy resolution to numerous emotion categories. Our work delves into the frontier of SER and attains greater insight into how deep learning may be leveraged to identify emotional states.