Samy Bengio Research Scientist in Machine Learning Google Inc 1600 Amphitheatre Parkway 94043 Mountain View California, USA Phone: +1 (650) 253 2563 Email: bengio [at] google.com

Publications and Vitae

• List of my publications.
• My Google Scholar profile.
• My curriculum vitae in postscript, pdf (acrobat) or html.
• The Statistical Machine Learning from Data lectures that I gave at EPFL.

Research Interests

• The Torch library
  The development of Torch (see also the more recent sourceforge link), a machine learning library in C++/GPL, including neural networks, support vector machines, hidden Markov models, mixture models, adaboost and bagging, etc. This library also includes the well-known SVMTorch software for support vector machines.
• Discriminant approaches to image and sound retrieval from text queries.
  How to efficiently build complex document (such as image or sound) retrieval systems from text queries that scale well and optimize a discriminant criterion.
• Discriminant approaches to speech processing
  While the orthodoxy in speech recognition over-uses hidden Markov generative models, large margin discrimant approaches are very interesting.
• A study of the difficulty of training deep architectures
• Biometric Authentication and Multimodal Fusion
• The Expected Performance Curve
  How to replace ROC curves to better estimate expected performance on unseen data.
• Cursive Handwriting Recognition

Former PhD Students/Postdocs/Research Assistants

• Mark Barnard (PhD 2005)
• Fabien Cardinaux (PhD 2005)
• Silvia Chiappa (PhD 2006)
• Ronan Collobert (PhD 2004)
• Christos Dimitrakakis (PhD 2006)
• Nicolas Gilardi (PhD 2002)
• David Grangier (PhD 2008)
• Mikaela Keller (PhD 2006)
• Johnny Mariéthoz (PhD 2006)
• Miguel Moreira (PhD 2001)
• Jean-Francois Paiement (PhD 2008)
• Norman Poh (PhD 2006)
• Alexei Pozdnoukhov (PhD 2006)
• Yann Rodriguez (PhD 2006)
• Conrad Sanderson (postdoc 2004)
• Katrin Weber (PhD 2003)
• Dong Zhang (PhD 2006)

Some Pointers

• We're organizing BigVision2012, a NIPS 2012 workshop on Big Data for Computer Vision.
• We're organizing WebVision2012, an ECCV 2012 workshop on Computer Vision for the Web.
• We're organizing BayLearn, a new workshop in machine learning for researchers in the Bay Area.
• A call for paper on Learning Deep Architectures organized by TPAMI in which I am a guest editor.
• A new book: Automatic Speech and Speaker Recognition: Large Margin and Kernel Methods, co-edited by Joseph Keshet and myself. Published by Wiley, 2009.
• The International Conference on Machine Learning (ICML), Montreal, QC, Canada, June 14-18, 2009.
• The NIPS Workshop on Efficient Machine Learning, Whistler, BC, Canada, December 7-8, 2007.
• The NIPS Workshop on Learning to Compare Examples, Whistler, BC, Canada, December 8-9, 2006.
• The MLMI'06 workshop, Washington DC, USA, May 2006.
• The MLMI'05 workshop, Edinburgh, UK, July 2005.
• The PASCAL network of excellence.
• The NIPS 2004 Workshop on Multimodal Signal Processing .
• The MLMI'04 workshop, Martigny, Switzerland, June 2004, and the corresponding MLMI'04 recordings of the talks.
• The 2002 IEEE Workshop on Neural Networks for Signal Processing (NNSP), Martigny, Switzerland, September 2002
• Description of the LISA laboratory (UdeM) .
• Description of the CIRANO research center.
• Description of the IDIAP Research Institute, where I worked for many years.

Varia

• My best Machine Learning experiments: Alice and Jeremy