An Overview of Bengali Speech Recognition: Methods, Challenges, and Future Direction
Authors:
- 8 March 2023
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 0873-0878
Abstract
In the subject of human-computer interactions, speech recognition is an appealing technique that gives users the opportunity to interact with and control the machine. Currently, automatic speech recognition (ASR) systems are being utilized to flawlessly convert speech to text. The implementation of ASR systems in Bengali has not yet achieved an acceptable standard, despite the fact that they are being used extensively in other languages. So far, various ASR models have been implemented for speech recognition like LSTM (Long Short-Term Memory), Transformer-based models like RNN (Recurrent Neural Network) and CNN (Convolutional Neural Network) are also quite popular for speech recognition. The Bengali language is more grammatically and structurally diverse than English. Therefore, it is difficult for researchers to use the same language model as English or any other language. So, the Bengali language is difficult to work with. Different studies have been carried out on Bengali speech recognition. We want to enlist the numerous models that have been used for Bengali speech recognition from 2009 to 2022. In this paper, we will discuss the challenges that were faced and the scope of future research in this field. This survey paper also provides datasets utilized in numerous research studies.Keywords
This publication has 26 references indexed in Scilit:
- A Novel Deep Learning based Sentiment Analysis of Twitter Data for US Airline Service2021
- Implementation Of Recurrent Neural Network With Language Model For Automatic Articulation Identification System In BanglaInternational Journal of Advanced Networking and Application, 2021
- Bangla Voice Command Recognition in end-to-end System Using Topic Modeling based Contextual Rescoring2020
- Bengali Spoken Digit Classification: A Deep Learning Approach Using Convolutional Neural NetworkProcedia Computer Science, 2020
- A Speech Recognition System for Bengali Language using Recurrent Neural Network2019
- Crowd-Sourced Speech Corpora for Javanese, Sundanese, Sinhala, Nepali, and Bangladeshi Bengali2018
- Bengali speech recognition: A double layered LSTM-RNN approach2017
- Acoustic modeling using deep belief network for Bangla speech recognition2015
- Developing Bengali Speech Corpus for Phone Recognizer Using Optimum Text Selection Technique2011
- Bangla Speech Recognition System Using LPC and ANN2009