Automatic gender classiﬁcation is a system to classify the gender using speech signal developed for speech encoding, analysis, synthesis and gender identiﬁcation. Generally generation of gender recognition system can be broadly in two levels namely front-end system and back-end system. The function of the front-end system is to represent by a set of vectors called feature such as pitch, short time energy (STE), energy entropy (EE), zero crossing rate (ZCR). The back-end system is also referred to classiﬁer, and it includes to develop a gender model to recognize the gender from speech signal. In our existing system uses fuzzy logic and neural network approach does not produced the exact required result for gender classiﬁcation due to complexity of training network. To overcome this problem various evolutionary algorithms like Genetic Algorithm (GA) is applied in gender classiﬁcation. This work here applies GA to select the features that are responsible for gender identiﬁcation. The implementation result shows the performance of the proposed method in gender identiﬁcation.
Digital Object Identifier (DOI)
Jayasankar, T.; Vinothkumar, K.; and Vijayaselvi, Arputha
"Automatic Gender Identiﬁcation in Speech Recognition by Genetic Algorithm,"
Applied Mathematics & Information Sciences: Vol. 11:
3, Article 31.
Available at: https://digitalcommons.aaru.edu.jo/amis/vol11/iss3/31