|
|
Hi.
I am
a 5th year Ph.D. student at Columbia University in the Laboratory for
the
Recognition and Organization of Speech and Audio, LabROSA,
under the supervision of professor Dan
Ellis. I also collaborate in diverse projects with Nebojsa
Jojic at Microsoft Research and with Bhiksha
Raj at Mitsubishi Electric Research Labs (MERL)
!!!
I MOVED !!!
I am now at Microsoft Research as a
post
doctoral researcher in the Machine Learning and Applied Statistics
Group,
MLAS.
I will be defending my thesis
sometime
in June. I can still be contacted at my Columbia address.
My almost none existant (for now)
new
website at MSR is: http://research.microsoft.com/~manuelrg
|
Courses
C.V.
|
Research
Interests
-
Pattern
Recognition and Machine Learning Techniques for Digital Audio Organization
and Modeling.
-
Audio
Source Separation.
-
Speech
Recognition
-
Content
Based Audio Classification and Retrieval(see
a review of (not so) recent papers)
|
Research
Projects
|
Publications
-
M. J. Reyes-Gomez, N. Jojic and D. P.W.
Ellis, ``Dominant
Speaker Segmentation of Sound Mixtures with the Deformable Spectrogram
Model''. In preparation for IEEE Transactions on Speech and Audio
Processing, February 2005
-
M. J. Reyes-Gomez, N. Jojic and D. P.W. Ellis
``The Deformable
Spectrogram Model for Sound Dynamics'' Website
Submitted to Special Issue on Statistical Models, IEEE Transactions on
Speech and Audio Processing, January 2005
-
M. J. Reyes-Gomez, N. Jojic and D. P.W.
Ellis. Deformable
Spectrograms Website
Artificial Intelligence and Statistics,
AISTATS'05. Barbados,
January 2005
-
M. J. Reyes-Gomez, N. Jojic and D. P.W.
Ellis. Towards
Single-Channel Unsupervised Source Separation of Speech Mixtures: The
Layered
Harmonics/Formants Separation/Tracking Model. Research Workshop on
Statistical and Perceptual Audio Processing, SAPA04, Korea October,
2004.
-
N. Jojic, Y. Caspi and M. J. Reyes-Gomez. Probabilistic
Index Maps for Modeling Natural Signals. Conference on Uncertainty
in Artificial Intelligence. Banff, Canada, July 2004
-
M.J.
Reyes-Gomez,N. Jojic and
D.P.W. Ellis. Detailed
graphical models for source separation and missing
data interpolation in audio.
2004 Snowbird Learning Workshop Snowbird, Utah, March 2004
-
M.J.
Reyes-Gomez, D.P.W. Ellis
and N. Jojic.
Multiband Audio Modeling for Single Channel Acoustic Source
Separation,
ICASSP Montreal 2004
-
B. Raj, M.L. Seltzer and M.J. Reyes-Gomez, "Speech
Recognizer Based Maximum Likelihood Beamforming", Book chapter,
Speech
Separation by Humans and Machines, ed. P. Divenyi, Kluwer.
-
M.J.
Reyes-Gomez, B. Raj and
D.P.W. Ellis. Multi-channel
Source Separation by Beamforming Trained with Factorial HMMS,
IEEE Workshop on Applications of Signal Procesing to
Audio and Acoustics,
New Paltz, NY 2003
-
M.J. Reyes-Gomez
and D.P.W Ellis(2003).
Selection,
Parameter Estimation and Discriminative Training of Hidden Markov
Models
for Generic Acoustic Modeling International
Conference
on Multimedia and Expo, Baltimore 2003
-
M.J.
Reyes-Gomez, B. Raj and
D.P.W. Ellis. Multi-channel
source separation by factorial HMMs, ICASSP Hong Kong
2003
-
M.J. Reyes-Gomez
and D.P.W.
Ellis. Error
visualization for tandem acoustic modeling on the Aurora task
ICASSP, Orlando, May 2002.
-
D.P.W. Ellis and
M.J. Reyes
Gomez (2001). Investigations
into Tandem Acoustic Modeling for the Aurora Task Proc.
Eurospeech-01,
Denmark, September 2001.
-
M. J. Reyes-Gomez. User-Defined
Object Classifier based in Neural Network with Optimal Selection of
Visual
Features using a Genetic Algorithm. Columbia University Technical
Report,
1999
|
Invited
Talks/ Posters
-
Manuel J.
Reyes-Gomez, Nebojsa
Jojic and Daniel P.W. Ellis. "Deformable Spectrograms" IBM
Research - Multimedia - Seminars, Hawthorne
NY, March 2005
-
Manuel J.
Reyes-Gomez, Nebojsa
Jojic and Daniel P.W. Ellis. "Deformable Spectrograms" National
Science Foundation Workshop on Speech Separation and Comprehension in
Complex
Acoustic Environments. Montreal Quebec November 2004
-
Manuel J.
Reyes-Gomez, Nebojsa
Jojic and Daniel P.W. Ellis. "Harmonics/Formants Separation and
Tracking
Model''. International Conference "Auditory Scene Analysis and
Speech
Perception by Human and Machine''. Hanse
Institute
for Advanced Study, Delmenhorst, Germany August 2004.
-
Manuel J.
Reyes-Gomez. "Statistical
Models for Source Separation, Scene Analysis and Other Audio
Applications''
Victoria University, Victoria Canada,
July 2004.
-
Manuel J.
Reyes-Gomez. "Graphical
Models and Variational Methods for Source Separation and other Audio
Applications.''
2003 IBM User Interface Technology
(UIT) Student Symposium.
IBM T.J. Watson Research Center, NY. September 2003
|