Multitask National Speech Corpus (MNSC v1) is derived from IMDA's NSC Corpus. MNSC is a multitask speech understanding dataset derived and further annotated from IMDA NSC Corpus. It focuses on the knowledge of Singapore's local accent, localised terms, and code switching. ASR: Automatic Speech Recognition SQA: Speech Question Answering SDS: Spoken Dialogue Summarization PQA: Paralinguistic Question Answering from datasets import load dataset data =… See the full description on the dataset page: https://huggingface.co/datasets/MERaLiON/Multitask National Speech Corpus v1.
We use cookies for essential functionality and analytics. You can accept or reject analytics cookies.Cookie policy