Workshop topics
Papers in all areas of distant-talking human/human and human/machine interaction are encouraged, including:
- Multi-channel and single-channel approaches for speech acquisition, noise suppression, source localization and separation, dereverberation, echo cancellation, and acoustic event detection
- Speech and speaker recognition technology for hands-free scenarios, including robust features, feature-domain enhancement and dereverberation, and model adaptation
- Microphone array technology and architectures, especially for distant-talking speech recognition and acoustic scene analysis
- Speech corpora for training and evaluation of distant-talking speech systems
- Applications based on microphone arrays and hands-free speech systems.
Paper & demo submission
The workshop technical program will consist of oral presentations, poster sessions, and demonstrations. Prospective authors are invited to submit full-length papers up to four pages, with a fifth page permitted for references only. Submissions for proposed demonstrations may be up to two pages in length.
Manuscripts should be prepared using the same format as for ICASSP submissions using the ICASSP author kit for LaTeX or Word.
Accepted papers will be published in IEEEXplore.
Special sessions
The workshop will feature two special sessions:
- Advances in sparse modeling and low-rank modeling for speech processing
proposed by Hervé Bourlard and Afsaneh Asaei (Idiap) - Speech detection and speaker localization in domestic environments
proposed by Maurizio Omologo (FBK) and the DIRHA consortium.
Best paper awards
The best paper and the best student paper will each receive a 500$ award sponsored by Starkey and IBM.
