An efficient feature selection method for arabic and english speech emotion recognition using Grey Wolf Optimizer

Nowadays, analyzing and interpreting emotions through human speech communication have drawn a great attention in the field of human-computer interaction. Therefore, many speech recognition systems have been suggested to recognize the emotional states of the speaker utilizing the speech recordings of their spoken utterances. Feature extraction is an important step in building an emotion recognition sys-tem in which it is used to extract emotional features from speech data. However, not all extracted fea-tures are relevant to classify the emotion states of the speaker. The existence of irrelevant and redundant features generates unmeaningful patterns that lead to inaccurate and undesirable emotion classification performance. Therefore, this study proposes an intelligent feature selection method based on a novel bio-inspired optimization algorithm that mimics the hunting mechanism of wolves in the nat-ure, called Grey Wolf Optimizer (GWO) and K-nearest neighbor (KNN) classifier, to find the most relevant subset of features to enhance the classification performance of an emotion recognition systems. The pro-posed method is called GWO-KNN. Emotion classification is performed on three distinct databases including Arabic Emirati-accented speech database, Ryerson Audio-Visual Database of Emotional Speech and Song dataset (RAVDESS), and Surrey Audio-Visual Expressed Emotion dataset (SAVEE). A combined or single feature extraction method is applied to extract the features from each dataset. The proposed method provides better classification performance for speech emotion recognition system com-pared to classical methods such as bat algorithm (BAT), cuckoo search (CS), White Shark Optimizer (WSH), and arithmetic optimization algorithm (AOA). Our proposed method also surpasses several state-of-the-art recent approaches that use the same datasets.(c) 2023 Elsevier Ltd. All rights reserved.

Automated summary

Nowadays, the analysis and interpretation of emotions in human speech communication have gained significant attention in the field of human-computer interaction. Various speech recognition systems have been proposed to recognize the emotional states of speakers through their speech recordings. Feature extraction is a critical step in building an emotion recognition system, but not all extracted features are relevant for classifying emotions accurately. This study introduces an intelligent feature selection method called GWO-KNN, which uses a bio-inspired optimization algorithm and a K-nearest neighbor classifier to enhance the classification performance of emotion recognition systems by identifying the most relevant subset of features. The proposed method outperforms classical methods and recent state-of-the-art approaches on three different databases.