The Austrian German Multi-Sensor Corpus
AMISCO is a collection of multi-room and multi-channel close- and distant-talking Austrian German high-quality speech-recordings from 24 speakers, balanioced male and female. It contains around 8.2 hours of read speech, 53,000 word tokens based on 2,070 unique word types. This corpus features glottograms, fundamental frequencies, positions, and video recordings of speakers located at certain positions or walking along trajectories provided by the Kinects’ skeleton tracker.
The whole corpus will be available from summer/autumn, 2017. Right now we are updating the corpus’s meta data, which will be soon available as a Git repository.
A preview is already available. Just click on the download-link below.
Short Descriptions of Preview
The downloadable preview contains
- audio data (48 kHz sampling frequency, 24 bit PCM S24 LE (araw) encoder),
- video data (30 fps),
- Kinect data (30 fps),
- extracted fundamental frequencies, as well as
- orthographic transcriptions.
We uploaded the preview for reviewers only. The whole corpus as well as manuals and further details will be available soon. For further information on the corpus, please refer to the submitted article. If the download fails, please contact Hannes Pessentheiner.
The corresponding (LREC) publication is already available: