The RedDots project is dedicated to the study of speaker recognition under conditions where test utterances are of short duration and of variable phonetic content. At the current stage, we focus on English speakers, both native and non-native, recruited worldwide. This is made possible through the use of a recording front-end consisting of an application running on mobile devices communicating with a centralized web server at the back-end. Speech recordings are collected by having speakers read text prompts displayed on the screen of the mobile devices. We aim to collect a large number of sessions from each speaker over a long time span, typically one session per week over a one year period. The corpus is expected to include rich inter-speaker and intra-speaker variations, both intrinsic and extrinsic (that is, due to recording channel and acoustic environment). The RedDots DatasetThe RedDots dataset was designed to have smaller number of recordings per session with more sessions of shorter duration spent on each. One targeted scenario is to collect 52 sessions per speaker, one session every week, for a year. In this regard, each session is limited to two minutes, where the number of sentences is determined to be 24 for each session. The composition of sentences used for a recording session is shown below.
We have released the RedDots dataset collected up to August 17 2015. Details of the current release can be found here.
In conjunction with the current release of the RedDots dataset, the first RedDots Challenge will be held within the scope of a special session at INTERSPEECH 2016. Please submit your paper to the special session "The RedDots Challenge: Toward Short-Term Characterization of Speaker" in INTERSPEECH 2016. More details about the RedDots Challenge are available here.
Relevant Information |
|
|