Mixer 6 - CHiME 8 Transcribed Calls and Interviews
| Item Name: | Mixer 6 - CHiME 8 Transcribed Calls and Interviews |
| Author(s): | Matthew Wiesner, Desh Raj, Matthew Maciejewski, Chloe Haviland, Samuele Cornell, Eleanor Chodroff, Sanjeev Khudanpur, Jack Godfrey |
| LDC Catalog No.: | LDC2025S07 |
| ISLRN: | 017-424-674-662-6 |
| DOI: | https://doi.org/10.35111/pk0y-qp29 |
| Release Date: | August 15, 2025 |
| Member Year(s): | 2025 |
| DCMI Type(s): | Sound, Text |
| Sample Type: | 16-bit FLAC |
| Sample Rate: | 16000 |
| Data Source(s): | microphone conversation, telephone conversations |
| Application(s): | speaker identification, speech recognition |
| Language(s): | English |
| Language ID(s): | eng |
| License(s): |
LDC User Agreement for Non-Members |
| Online Documentation: | LDC2025S07 Documents |
| Licensing Instructions: | Subscription & Standard Members, and Non-Members |
| Citation: | Wiesner, Matthew, et al. Mixer 6 - CHiME 8 Transcribed Calls and Interviews LDC2025S07. Web Download. Philadelphia: Linguistic Data Consortium, 2025. |
| Related Works: | View |
Introduction
Mixer 6 - CHiME 8 Transcribed Calls and Interviews was developed for the 7th and 8th CHiME (Computational Hearing in Multisource Environments) challenges. It contains 80 hours of English interviews and telephone speech from Mixer 6 Speech (LDC2013S03) with transcripts developed for the CHiME challenges and divided into training, development and test sets. This data was used in CHiME 7 Task 1 and CHiME 8 Task 1 both of which focused on transcription and segmentation across varied recording conditions such as interviews, meetings, and dinner parties, with an emphasis on generalization across recording device types and array topologies.
Mixer 6 Speech was developed by the Linguistic Data Consortium (LDC) and comprises 15,863 hours of audio recordings of interviews, transcript readings and conversational telephone speech involving 594 distinct native English speakers recorded over 14 channels. This material was collected by LDC in 2009 and 2010 as part of the Mixer project, specifically phase 6, the focus of which was on native American English speakers local to the Philadelphia area.
Data
The data includes audio from Mixer 6 Speech recorded on 13 microphones for a total of 1063 hours corresponding to 80 hours of speech. The development and test splits are speaker-disjoint from the training data and consist of fully transcribed, multi-microphone interviews. The transcripts were developed in three phases: (1) manual transcription, segmentation and automatic alignment with speech; (2) splitting sessions into sets; and (3) splitting certain sessions from the training set. Each segment was labeled with the speaker, the uttered text, and the start and end times in seconds for that segment.
Audio data is provided as 16 bit FLAC files sampled at 16kHz. Transcripts are released as UTF-8 encoded JSON files.
Samples
Please view the following samples:
Updates
No updates at this time.