dataset_info:
config_name: CC_BY_3.0
features:
- name: text
dtype: string
- name: start
dtype: float64
- name: end
dtype: float64
- name: speaker
dtype: string
- name: language
dtype: string
- name: dnsmos
dtype: float64
- name: source_podcast
dtype: string
- name: audio
dtype: audio
- name: speaker_id
dtype: string
splits:
- name: train
num_bytes: 1437253098.316
num_examples: 17942
download_size: 1432758259
dataset_size: 1437253098.316
configs:
- config_name: CC_BY_3.0
data_files:
- split: train
path: CC_BY_3.0/train-*
license: cc
This particular dataset only kept the CC-BY 3.0 podcasts, which have been processed using the Emilia-Pipe with Whisper Large v3.
Some Podcasts
Podcasts are taken from the PodcastFillers dataset. The PodcastFillers dataset consists of 199 full-length podcast episodes in English with manually annotated filler words and automatically generated transcripts. The podcast audio recordings, sourced from SoundCloud, are CC-licensed, gender-balanced, and total 145 hours of audio from over 350 speakers.
This dataset doesn't upload the PodcastFillers annotations, which are under a non-commercial license. See here for more details.
Length by license type
CC_BY 3.0: Total length: 51.44h
License
See here for more details. The licenses are also in the metadata.
Citation Information
@inproceedings{Zhu:FillerWords:INTERSPEECH:22,
title = {Filler Word Detection and Classification: A Dataset and Benchmark},
booktitle = {23rd Annual Cong.~of the Int.~Speech Communication Association (INTERSPEECH)},
address = {Incheon, Korea},
month = {Sep.},
url = {https://arxiv.org/abs/2203.15135},
author = {Zhu, Ge and Caceres, Juan-Pablo and Salamon, Justin},
year = {2022},
}
Contributions
Thanks to @ylacombe for adding this dataset.