Spoken Language Identification using Deep Learning

A crucial problem in natural language processing is language identification, which has applications in speech recognition, translation services, and multilingual content. The five main Indian languages that are the subject of this study are Hindi, Bengali, Tamil, English, and Gujarati. A Deep Neural...

Full description

Saved in:
Bibliographic Details
Published in2024 International Conference on Electrical Electronics and Computing Technologies (ICEECT) Vol. 1; pp. 1 - 7
Main Authors Julius, Christian Anthony, Vijayalakshmi, S, Palathara, Tiny S
Format Conference Proceeding
LanguageEnglish
Published IEEE 29.08.2024
Subjects
Adaptation models
Artificial neural networks
Cepstral analysis
Data models
deep neural network
Language identification
Linguistics
major indian spoken languages
mel-frequency cepstral coefficients
Natural language processing
Robustness
spoken language
Transfer learning
Transformers
Virtual assistants
Online AccessGet full text
DOI10.1109/ICEECT61758.2024.10738935

Cover

More Information
Summary:A crucial problem in natural language processing is language identification, which has applications in speech recognition, translation services, and multilingual content. The five main Indian languages that are the subject of this study are Hindi, Bengali, Tamil, English, and Gujarati. A Deep Neural Network is introduced in the paper which is specifically made to use Mel-Frequency Cepstral Coefficients (MFCCs) for sophisticated language categorization. The suggested architecture of the model, which includes batch normalisation and tightly linked layers, helps it to be adept at identifying complex linguistic patterns. Comparing the research to the source work [18], promising improvements are shown, highlighting the potential of the model in language detection.
DOI:10.1109/ICEECT61758.2024.10738935