Philippe Hamel

Français

Philippe Hamel Name: Philippe Hamel
email : hamelphi ( A T ) iro (dot) umontreal (dot) ca
Ph.D Candidate
Département d'informatique et de recherche opérationelle (DIRO)
Universtié de Montréal
Member of:
LISA
GAMME
BRAMS
CIRMMT

Hello,

I am a Ph.D student in computer science at the Université de Montréal. I am studying machine learning and its applications to music.

Ph. D. Publications (Computer Science):

Philippe Hamel, Simon Lemieux, Yoshua Bengio and Douglas Eck, Temporal pooling and multiscale learning for automatic annotation and ranking of music audio, In Proceedings of the 12th International Conference on Music Information Retrieval (ISMIR), Miami, FL, USA, 2011

Jason Weston, Samy Bengio and Philippe Hamel, Multi-Tasking with Joint Semantic Spaces for Large-Scale Music Annotation and Retrieval, Journal of New Music Research (2011)

Philippe Hamel, Pooled features classifier MIREX 2011 contest entry. Best performance in audio tagging task.

Philippe Hamel and Douglas Eck, Learning features from music audio with deep belief networks, In Proceedings of the 11th International Society for Music Information Retrieval Conference (ISMIR), Utrecht, The Netherlands, pages 339--344, 2010

Philippe Hamel, Sean Wood and Douglas Eck, Automatic Identification of Instrument Classes in Polyphonic and Poly-Instrument Audio, in 10th International Society for Music Information Retrieval Conference (ISMIR), Kobe, Japan, pages 399--404, 2009

M. Sc. Publications (Theoretical physics) :

P. Hamel, "La nouvelle physique dans le système des mésons B", Master’s thesis, Université de Montréal, 2006.

S. Baek, A. Datta, P. Hamel, O. Hernández, and D. London, "Polarization states in B→ρK* and new physics", Physical Review D, vol. 72, 2005, p. 20.

S. Baek, P. Hamel, D. London, A. Datta, and D. Suprun, "B→πK puzzle and new physics", Physical Review D, vol. 71, 2005, p. 4.


Research Interests

I am studying machine learning, a sub-field of artificial intelligence applied to music. 'What is machine learning ?' you may ask. Well, in short, machine learning is the art of developping algorithms that can learn by examples. It is trying to recognize patterns and pull out the important information from a large quantity of data, and then be able to make good predictions over new examples. That is callled generalization and it is the key to good machine learning. Click here for a better description of machine learning.

Machine learning can be applied to many fields such as computer vision, hand-written character recognition or speech recognition. I am studying the application of machine learning to the domain of music. Because music is abstract and complex, it is really hard for machines to understand music. Machine learning can help to solve tasks such as genre recognition, source separation, beat tracking, optical music recognition, instrument recognition and many more tasks. I am particulary interested in finding ways to represent music in ways that are useful and meaningful for algorithms and humans.


Musical Interests
I am an amateur musician. I love to jam with friends, and am pretty open about musical styles. My main influences would be artists such as Pink Floyd, Radiohead, Godspeed You Black Emperor, Jean Leloup, Karkwa, Patrick Watson, The Doors and many more...

My main instrument is the guitar(s), but I like to make music (or noise) with all the instruments I can try. I can also play piano, drums, bass, harmonica, etc.

General interests
I am interested in sciences in general, but have a particular interest for physics. I am curious about about how nature works, and how much we can know about the universe.


Education

 
Ph.D. Candidate in Computer Science (Begun May 2008)
Université de Montréal (FQRNT research scolarship )
 
M. Sc. Physics (April 2006) Theoretical particle physics
Université de Montréal  (FQRNT research scolarship)
 
B.Sc in Physics and Mathematics (April 2004)
Université de Montréal

Last updated October 31 2011