Hugging Face's logo

Datasets:
PalashJain
/
VITB-HEBiC_Code-switch_corpus

Tasks:
Automatic Speech Recognition
Modalities:
Audio
Formats:
soundfolder
Languages:
English
Hindi
Size:
< 1K
Libraries:
Datasets
Dataset card Viewer Files Community
1

You need to agree to share your contact information to access this dataset

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Log in or Sign Up to review the conditions and access this dataset content.

Dataset Summary

VITB-HEBiC (VIT Bhopal Hindi-English bilingual corpus) is a Hindi-English code-switching corpus created at Vellore Bhopal University. This corpus consists of 7.5 hours of read speech (audio files) along with text and has 58,245 words. The corpus is highly diverse and includes 137 speakers from 27 states of India, each exhibiting distinct accents and dialects. All recordings are in natural environment and in .wav format with sampling rate of 16KHz, 16-bit PCM encoded and bit rate of 256kbps.

Downloads last month
9
Edit dataset card