SpeechOcean762
speechocean762
speechocean762: A non-native English corpus for pronunciation scoring task
Introduction
Pronunciation scoring is a crucial technology in computer-assisted language learning (CALL) systems. The pronunciation quality scores might be given at phoneme-level, word-level, and sentence-level for a typical pronunciation scoring task.
This corpus aims to provide a free public dataset for the pronunciation scoring task.
Key features:
It is available for free download for both… See the full description on the dataset page: https://huggingface.co/datasets/mispeech/speechocean762.speechocean762-librispeech-posteriors
SpeechOcean762 LibriSpeech Acoustic Posterior Exports
This repository stores frame-level acoustic-model output exports for the
SpeechOcean762 train and test sets, computed with LibriSpeech-trained acoustic
models and Charsiu models.
Files
test/<model_id>/posteriors.h5
contains one model's padded tensor for the test split.
train/<model_id>/posteriors.h5
contains one model's padded tensor for the train split.
manifests/train.json and manifests/test.json
are the… See the full description on the dataset page: https://huggingface.co/datasets/Haopeng/speechocean762-librispeech-posteriors.speechocean762
speechocean762: A non-native English corpus for pronunciation scoring task
How to use?
you can load data using
speechocean762_dataset = load_dataset('seba3y/speechocean762')
>> speechocean762_dataset
DatasetDict({
train: Dataset({
features: ['spk', 'age', 'gender', 'utt_name', 'audio', 'utt_text', 'utt_accuracy', 'utt_completeness', 'utt_fluency', 'utt_prosodic', 'utt_total', 'words', 'words_accuracy', 'words_stress', 'words_total', 'phones'… See the full description on the dataset page: https://huggingface.co/datasets/seba3y/speechocean762.speechocean762
Dataset Card for "speechocean762"
The datasets introduced in
Zhang, Junbo, et al. "speechocean762: An open-source non-native english speech corpus for pronunciation assessment." arXiv preprint arXiv:2104.01378 (2021).
Currently, phonetic-level evaluation is omitted (total sentence-level scores are just used.)
The original full data link: https://github.com/jimbozhang/speechocean762
More Information needed
speechocean762_preprocessed_for_whisperspeechocean762_fluency