2012 NIST Speaker Recognition Evaluation Test Set
| Item Name: | 2012 NIST Speaker Recognition Evaluation Test Set |
| Author(s): | Craig Greenberg, Alvin Martin, David Graff, Kevin Walker, Stephanie Strassel, Karen Jones |
| LDC Catalog No.: | LDC2026S09 |
| ISLRN: | 738-919-827-208-6 |
| DOI: | https://doi.org/10.35111/q6k0-mg59 |
| Release Date: | July 15, 2026 |
| Member Year(s): | 2026 |
| DCMI Type(s): | Sound |
| Sample Type: | sph (pcm, mulaw) |
| Sample Rate: | 8000, 16000 |
| Data Source(s): | microphone conversation, microphone speech, telephone conversations |
| Project(s): | NIST SRE |
| Application(s): | speaker identification |
| Language(s): | English |
| Language ID(s): | eng |
| License(s): |
LDC User Agreement for Non-Members |
| Online Documentation: | LDC2026S09 Documents |
| Licensing Instructions: | Subscription & Standard Members, and Non-Members |
| Citation: | Greenberg, Craig, et al. 2012 NIST Speaker Recognition Evaluation Test Set LDC2026S09. Web Download. Philadelphia: Linguistic Data Consortium, 2026. |
| Related Works: | View |
Introduction
2012 NIST Speaker Recognition Evaluation Test Set was developed by the Linguistic Data Consortium (LDC) and NIST (National Institute of Standards and Technology). It contains approximately 10,321 hours of English conversational telephone speech and in-person recorded studio sessions for evaluation and modeling, along with answer keys, trial files and documentation from the NIST-sponsored 2012 Speaker Recognition Evaluation (SRE).
The ongoing series of SRE evaluations conducted by NIST are intended to support research in text-independent speaker recognition. To this end, the evaluations are designed to be simple, to focus on core technology issues, to be fully supported, and to be accessible to researchers wishing to participate.
The SRE task is speaker detection, that is, determining whether a specified target speaker is speaking in a given segment of speech. SRE12 introduced a revised evaluation structure in which training data for target speakers was drawn from prior SRE corpora developed by LDC and was provided in advance of the evaluation period. Further information about the evaluation is contained in the SRE12 evaluation plan included in this release.
Data
Test data was drawn from Mixer 7 English Speech (LDC2025S08) and REMIX Telephone Collection (LDC2023S09). Those datasets also provided segments for modeling data; other modeling segments were drawn from Mixer 3 Speech (LDC2023S02), Mixer 4 and 5 Speech (LDC2020S03), and Mixer 6 Speech (LDC2013S03). The Mixer telephone protocol connected recruited speakers through a robot operator to carry on casual conversations; Mixer participants also completed on-site telephone calls and interview sessions conducted on a cross-channel platform. The test data contains English speech only; some non-English speech is contained in modeling segments.
This release is comprised of 130,844 test segments (6,623.01 hours), specifically, 3,946 hours of telephone speech (83,778 call segments) and 2,677 hours of interview speech (47,066 interview segments). Modeling data consists of 46,948 segments (3,704.65 hours).
Audio data is presented as uncompressed NIST SPHERE files. In-room microphone recordings are provided as 2-channel, 16-bit PCM, 16 kHz audio, and telephone recordings are provided as 8-bit Mu-law, 8 kHz audio, including both summed single-channel and separate-channel formats.
This corpus also contains answer keys, trial files and documentation.
Samples
Updates
No updates at this time.