YOHO Speaker Verification

The YOHO database contains a large scale, high-quality speech corpus to support text-dependent speaker authentication research, such as is used in secure access technology. The data was collected in 1989 by ITT under a US Government contract, but has not been available for public use before. Note that certain changes have been made to the corpus, mainly to insure the privacy of the speakers and some data has been withheld by the government for future use in testing. <p> YOHO contains: </p><ul> <li> Combination lock phrases (e.g. 36-24-36) </li> <li> Collected over three-month period in a real-world office environment </li> <li> Four enrollment sessions per subject with 24 phrases per session </li> <li> Ten test sessions per subject with four phrases per session </li> <li> 8kHz sampling with 3.8 kHz analog bandwidth </li> <li> 1.5 gigabytes of data </li> </ul> The number of trials is thus sufficient to permit evaluation testing at high confidence levels. In each session, a speaker was prompted with a series of phrases to be read aloud each phrase was a sequence of three two-digit numbers (e.g. 35 - 72 - 41, pronounced thirty-five seventy-two forty-one). The first four sessions for a given speaker were enrollment sessions of 24 phrases and all additional sessions were verification trials of four phrases each. In all there are 552 enrollment sessions and 1,380 trial sessions, with a nominal time interval of three days between sessions.<h3>Updates</h3> <p> An <a href="ftp://ftp.ldc.upenn.edu/pub/ldc/public_data/YOHO_update.tar" rel="nofollow">update</a> is available that corrects a bug in the original release.</p> </br>