Audio
/
Text
Japanese Two-Speaker LR-Separated Private Dialogue Speech with Transcripts
A dataset of private dialogue speech and transcripts featuring two speakers, with each channel recorded separately.
Subject Attributes
Japanese individuals (87 pairs) with diverse genders and ages
Data Format
mp3
License
Commercial and research use permitted / Rights-cleared / Free for academic use (details here)
Notes
[Recording Duration] 500 hours (approx. 60 min per file)
[Recording Environment] Web conference, etc. / Stereo LR separation
[Audio Rate] 48kHz / 192kbps
[Dialogue Content] Conversation scenes about hobbies, special skills, values, etc.
Sample
* Samples of this dataset are available only to those who apply through the form.
Request a Free Sample
