Sefara, Tshephisho JMokgonyane, TB2021-10-072021-10-072021-08Sefara, T.J. & Mokgonyane, T. 2021. Gender identification in Sepedi speech corpus. http://hdl.handle.net/10204/12120 .978-1-7281-8592-7978-1-7281-8591-0978-1-7281-8593-4DOI: 10.1109/icABCD51485.2021.9519308http://hdl.handle.net/10204/12120Gender identification is the task of identifying the gender of the speaker from the audio signal. Most gender identification systems are developed using datasets belonging to well-resourced languages. There has been little focus on creating gender identification systems for under resourced African languages. This paper presents the development of a gender identification system using a Sepedi speech dataset containing a duration of 55.7 hours made of 30776 males and 28337 females. We build a gender identification system using machine learning models that are trained using multilayer Perceptron (MLP), convolutional neural network (CNN), and long short-term memory (LSTM). Mid-term features are extracted from time domain features, frequency domain features and cepstral domain features, and normalised using the Z-score normalisation technique. XGBoost is used as a feature selection method to select important features. MLP achieved the same F-score and an accuracy of 94% for data with seen speakers while LSTM and CNN achieved the same F-score and an accuracy of 97%. We further evaluated the models on data with unseen speakers. All the models achieved good performance in F-score and accuracy.FulltextenGender identificationConvolutional neural networkSepediXGBoostFeature selectionLong short-term memoryMultilayer PerceptronConference PresentationSefara, T. J., & Mokgonyane, T. (2021). Gender identification in Sepedi speech corpus. http://hdl.handle.net/10204/12120Sefara, Tshephisho J, and TB Mokgonyane. "Gender identification in Sepedi speech corpus." <i>2021 International Conference on Artificial Intelligence, Big Data, Computing and Data Communication Systems (icABCD), Durban, South Africa, 5-6 August 2021</i> (2021): http://hdl.handle.net/10204/12120Sefara TJ, Mokgonyane T, Gender identification in Sepedi speech corpus; 2021. http://hdl.handle.net/10204/12120 .TY - Conference Presentation AU - Sefara, Tshephisho J AU - Mokgonyane, TB AB - Gender identification is the task of identifying the gender of the speaker from the audio signal. Most gender identification systems are developed using datasets belonging to well-resourced languages. There has been little focus on creating gender identification systems for under resourced African languages. This paper presents the development of a gender identification system using a Sepedi speech dataset containing a duration of 55.7 hours made of 30776 males and 28337 females. We build a gender identification system using machine learning models that are trained using multilayer Perceptron (MLP), convolutional neural network (CNN), and long short-term memory (LSTM). Mid-term features are extracted from time domain features, frequency domain features and cepstral domain features, and normalised using the Z-score normalisation technique. XGBoost is used as a feature selection method to select important features. MLP achieved the same F-score and an accuracy of 94% for data with seen speakers while LSTM and CNN achieved the same F-score and an accuracy of 97%. We further evaluated the models on data with unseen speakers. All the models achieved good performance in F-score and accuracy. DA - 2021-08 DB - ResearchSpace DP - CSIR J1 - 2021 International Conference on Artificial Intelligence, Big Data, Computing and Data Communication Systems (icABCD), Durban, South Africa, 5-6 August 2021 KW - Gender identification KW - Convolutional neural network KW - Sepedi KW - XGBoost KW - Feature selection KW - Long short-term memory KW - Multilayer Perceptron LK - https://researchspace.csir.co.za PY - 2021 SM - 978-1-7281-8592-7 SM - 978-1-7281-8591-0 SM - 978-1-7281-8593-4 T1 - Gender identification in Sepedi speech corpus TI - Gender identification in Sepedi speech corpus UR - http://hdl.handle.net/10204/12120 ER -24961