Workshop on machine hearing and learning

A workshop presented in collaboration with Research Axis 1 (Instruments, devices and systems).

     machine hearing image

 

Registration

Space is limited; ensure your seat by visiting Registration - Workshop on machine hearing and learning

Description

Machines are gradually becoming more adept in solving human perceptual tasks. This workshop aims to explore current and cutting-edge research, concepts, and workflows related to machine hearing and learning, auditory modelling, music and audio information retrieval, and audio (aural) feature extraction. Graduate students and invited participants will present their research for interdisciplinary discussion. 

Coffee and snacks will be provided. 

Schedule

 The complete schedule with affiliations and abstracts can be viewed in this pdf document.

  • 9:00-9:20 - Morning coffee and welcome
  • 9:20-9:45 - Rachel Bouserhal, Philippe Chabot, Milton Sarria-Paja, Patrick Cardinal, Jérémie Voix: Classification of nonverbal human produced audio events: A pilot study
  • 9:45-10:10 - Ajin Tom, Joshua Reiss, Philippe Depalle: An automatic mixing system for multitrack spatialization for stereo
  • 10:10-10:35 - Richard Lyon: Break the log jam: Per-channel energy normalization improves mel filterbank feature extraction for speech, music, bioacoustics, and sound events
  • 10:35-11:00 - Nicolas Figueiredo, Felipe Felix, Carolina Medeiros, Marcelo Queiroz: A comparative study on filtering and classification of bird songs
  • 11:00-11:15 - Coffee break
  • 11:15-11:40 - Dave Benson, Wieslaw Woszczyk: Predicting verbal descriptions of algorithmic reverb presets using audio signal features
  • 11:40-12:05 - Shayenne Moura, Marcelo Queiroz: Melody transcription system using deep neural networks
  • 12:05-12:30 - François Grondin, François Michaud: Lightweight sound source localization, tracking and separation methods for microphone array with arbitrary geometry 
  • 12:30-12:55 - Philippe Gournay, Reza LotfiDereshgi, Roch Levebvre: Artificial intelligence for speech and audio compression
  • 12:55-13:00 - Closing remarks

Call for Contributions (now closed)

Potential presenters are invited to send in their proposals to philippe-aubert.gauthier [AT] usherbrooke.ca. Each participant will be given 20 minutes, or less, depending on the number of proposals accepted. Short, rapid-fire presentations are also encouraged. 

 

 Submissions should include:

- author(s) name and affiliation;

- title;

- abstract (limited to 100 words); 

 

The deadline for submission of the proposals is 12:00 pm (noon), September 11, 2018.