Bengali
Emerging6papers using it
2023first seen
The 'Bengali' dataset is used to evaluate acoustic unit discovery and voice conversion tasks in the context of Textless NLP for Indian languages.
Papers using Bengali (6)
- Factors affecting ASR performance: A study using state of the art ASR models in Indic LanguagesMind the Pause: Disfluency-Aware Objective Tuning for Multilingual Speech Correction with LLMsModel Adaptation for ASR in low-resource Indian LanguagesAdversarial Training For Low-Resource Disfluency CorrectionPredicting positive transfer for improved low-resource speech
recognition using acoustic pseudo-tokensTextless NLP -- Zero Resource Challenge with Low Resource Compute