Rabindra Nath Nandi
Research Subjects
Rabindra Nath Nandi has a total of 9 co-authors. Here are the top co-authors by number of shared publications:
Firoj Alam
Qatar Computing Research Institute HBKU Doha Qatar
2 shared publicationsQuazi Sarwar Muhtaseem
Hishab Singapore Pte. Ltd Singapore
2 shared publicationsSagor Sarker
Hishab Singapore Pte. Ltd Singapore
2 shared publications- Shammur Absar Chowdhury
Qatar Computing Research Institute HBKU Doha Qatar
2 shared publications Tariqul Islam
2 shared publicationsMehadi Hasan Menon
Hishab Singapore Pte. Ltd Singapore
1 shared publication
Latest Publications
- Open Access
Pseudo-Labeling for Domain-Agnostic Bangla Automatic Speech Recognition
One of the major challenges for developing automatic speech recognition (ASR) for low-resource languages is the limited access to labeled data with domain-specific variations. In this study, we propose a pseudo-labeling approach to develop a large-scale domain-agnostic ASR dataset. With the proposed methodology, we developed a 20k+ hours labeled Bangla speech dataset covering diverse topics, speaking styles, dialects, noisy environments, and conversational scenarios. We then exploited the developed corpus to design a conformer-based ASR system. We benchmarked the trained ASR with publicly available datasets and compared it with other available models. To investigate the efficacy, we designed and developed a human-annotated domain-agnostic test set composed of news, telephony, and conversational data among others. Our results demonstrate the efficacy of the model trained on psuedo-label data for the designed test-set along with publicly-available Bangla datasets. The experimental resources will be publicly available.(https://github.com/hishab-nlp/Pseudo-Labeling-for-Domain-Agnostic-Bangla-ASR)
Open AccessPseudo-Labeling for Domain-Agnostic Bangla Automatic Speech Recognition
One of the major challenges for developing automatic speech recognition (ASR) for low-resource languages is the limited access to labeled data with domain-specific variations. In this study, we propose a pseudo-labeling approach to develop a large-scale domain-agnostic ASR dataset. With the proposed methodology, we developed a 20k+ hours labeled Bangla speech dataset covering diverse topics, speaking styles, dialects, noisy environments, and conversational scenarios. We then exploited the developed corpus to design a conformer-based ASR system. We benchmarked the trained ASR with publicly available datasets and compared it with other available models. To investigate the efficacy, we designed and developed a human-annotated domain-agnostic test set composed of news, telephony, and conversational data among others. Our results demonstrate the efficacy of the model trained on psuedo-label data for the designed test-set along with publicly-available Bangla datasets. The experimental resources will be publicly available.(https://github.com/hishab-nlp/Pseudo-Labeling-for-Domain-Agnostic-Bangla-ASR)
- Open Access
- 1 January 2023
- Published by Association for Computational Linguistics (ACL)
- p. 152-162
Pseudo-Labeling for Domain-Agnostic Bangla Automatic Speech Recognition
Rabindra Nath Nandi, Mehadi Menon, Tareq Muntasir, Sagor Sarker, Quazi Sarwar Muhtaseem, Md. Tariqul Islam, Shammur Chowdhury, Firoj Alam. Proceedings of the First Workshop on Bangla Language Processing (BLP-2023). 2023.
- 1 January 2023
- Published by Association for Computational Linguistics (ACL)
- p. 152-162
Open AccessPseudo-Labeling for Domain-Agnostic Bangla Automatic Speech Recognition
Rabindra Nath Nandi, Mehadi Menon, Tareq Muntasir, Sagor Sarker, Quazi Sarwar Muhtaseem, Md. Tariqul Islam, Shammur Chowdhury, Firoj Alam. Proceedings of the First Workshop on Bangla Language Processing (BLP-2023). 2023.