2026-05-12T20:39:49.312Z http://ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai

ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai:2:5085 2019-06-01T14:56:43Z 2

A Study of Language and Classifier-independentFeature Analysis for Vocal Emotion Recognition 2017 http://ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai/record/2/5085 1811.08935 F. Noroozi М. Марјановић-Јаковљевић A. NJeguš S. Escalera G. Anbarjafari Audio signals are commonly used for human-computer interaction. It is based on linguistic abilities and emotional statements. Every speech signal carries implicit information about the emotions, which can be extracted by speech processing methods. An important step in every automatic vocal based emotion recognition system is to select proper feature sets used for the classification of emotions. A robust emotion recognition system should overcome different languages either different methods boundaries. In this paper, we propose an algorithm for extracting features that are independent from the spoken language and the classification method to have comparatively good recognition performance on different languages independent from employed classification methods. The proposed algorithm is composed of three stages. In the first stage, we propose feature ranking method analyzing the state-of-the-art voice quality features. In the second stage, we propose a method for finding the subset of the common features for each language and classifier. In the third stage, we compare our approach with the recognition rate of the state-of-the-art filter methods. We use four databases with different languages, namely, Polish, Serbian, English and Italian. Also four different classifiers, namely, nearest neighbor, support vector machine and gradient descent Neural Network and Convolutional Neural Network, are employed. article 1 20 1811.0893 https://arxiv.org/abs/1811.08935