Dataset Specifics and Reading Material

Implementation details, file formats, and paper links

Dataset Description

The Voices Obscured in Complex Environmental settings (VOiCES) corpus presents audio recorded in acoustically challenging conditions. Recordings took place in real rooms of various sizes, capturing different background and reverberation profiles for each room. Various types of distractor noise (TV, music, or babble) were simultaneously played with clean speech. Audio was recorded at a distance using twelve microphones strategically placed throughout the room. To imitate human behavior during conversation, the foreground speaker used a motorized platform, rotating over a range of angles during recordings.

Three hundred distinct speakers from LibriSpeech’s “clean” data subset were selected as the source audio, ensuring a 50-50 female-male split. In preparation for upcoming data challenges, the first release of the VOiCES corpus will include 200 speakers only. The remaining 100 speakers will be reserved for model validation; the full corpus (300 speakers) will be released once the data challenge is closed.

In addition to the full dataset, we also provide a dev set and a mini-dev set. Both maintain the data structure of the VOiCES corpus, but include a small subset of data. The dev set includes audio files for four randomly selected speakers (50-50 female-male split) for data recorded in Room-1. This includes data from all twelve microphones. The mini-dev set includes one speaker, one room (Room-1), and studio microphones only.


For more details about how to use the dataset, please see our README.

Blog Posts