The IberSPEECH 2020 Challenge starts!
Zaragoza, March 1, 2020
Speaker Diarization is a very important task for some speech technologies like Automatic Speech Recognition, Speaker Identiﬁcation or Spoken Document Retrieval. This evaluation consists of segmenting broadcast audio documents according to different speakers and linking those segments which originate from the same speaker. No a priori knowledge is provided about the number or the identity of the speakers participating in the audio to be analyzed. This year an optional speaker identification step has been included in the evaluation. Given a closed set of characters the challenge will be to label the segments given by the speaker diarization step with the known character names.The calendar for the Albayzín Speaker Diarization and Identification evaluation is:
- March 23th, 2020: Registration opens
- March 23th, 2020: Release of the training and development data
June 1st, 2020September 7th, 2020: Registration deadline. Release of the evaluation data June 30th, 2020October 9th, 2020: Submission deadline July 15th, 2020October 30th, 2020: Results distribution to the participants September 13, 2020December 23th, 2020: Paper submission deadline November 18-20, 2020March, 2021: Iberspeech 2020 conference in Valladolid