German in-the-wild speech (talk shows, interviews, podcasts, vlogs and other YouTube content) paired with its transcript, acoustic measurements, discrete German descriptor tags and, for female speakers, a free-text German description of the voice and recording conditions. The dataset is intended for training description-conditioned TTS models such as The data was built as part of research work. It is a random subset of the German part of Emilia-YODAS, processed until the female split reached about 350 hours. Both splits have the same columns; only the female split is described. The split a clip lands in is decided by a gender classifier (see Columns), not by human labels, so a small share…
Access requested at publisher
cc-by-4.0
100K<n<1M
German European Parliament plenary speech paired with a transcript, acoustic measurements, discrete German descriptor tags and a free-text German description of the voice and recording conditions. The dataset is intended for training description-conditioned TTS models such as The data was built as part of research work. It is a random subset of the German unlabelled part of VoxPopuli (release v1, from the VoxPopuli GitHub repository), not the transcribed subset on the Hub (facebook/voxpopuli), which it does not overlap (see below). Most of the audio is simultaneous interpretation. VoxPopuli's German unlabelled data is the Parliament's German audio channel: when a member speaks German it…
Access requested at publisher
cc-by-4.0
100K<n<1M
