Speakers
The VoxPopuli dataset includes speech from multiple speakers, but how many speakers are represented in the dataset?