Amit Gruber

Postdoctoral Fellow
Department of Computer Science
University of Toronto
Office: Pratt 290F, 6 King's College Road, Toronto, ON.



Phone:  (416) 978 7391

I'm a postdoctoral fellow in computer science with Rich Zemel at the University of Toronto.
Previously, I was a PhD student of Yair Weiss at the School of Computer Science and Engineering at the Hebrew University of Jerusalem. My research interests are in the fields of Machine Learning and Computer Vision.

Open Source Software:

Open HTMM
A c++ open source implementation of EM inference in the Hidden Topic Markov Model.
I wrote this code at Google while working with Ashok Popat.

LTHM source code
A c++ open source implementation of EM inference in the Latent Topic Model for Hypertext.
The current implementation is the research code, built on top of the HTMM implementation with epsilon set to 1. It will soon be replaced with a cleaner and more efficient implementation.

Publications:

"Topic Models for Hypertext: How Many Words is a Single Link Worth?",
Amit Gruber, Michal Rosen-Zvi and Yair Weiss,
NIPS workshop on Analyzing Graphs: Theory and Applications, Whistler, Canada, December 2008.

"Latent Topic Models for Hypertext",
Amit Gruber, Michal Rosen-Zvi and Yair Weiss,
Uncertainty in Artificial Intelligence (UAI), Helsinki, Finland, July 2008.

"Sparse Regression as a Sparse Eigenvalue problem",
Baback Moghaddam, Amit Gruber, Yair Weiss and Shai Avidan,
Information Theory and Applications Workshop (ITA'08), January 2008.

"Incorporating non-motion cues into 3D motion segmentation",
Amit Gruber and Yair Weiss,
Computer Vision and Image Understanding (CVIU), Volume 108, Issue 3, December 2007, Pages 261-271, Special issue on Spatial Coherence in Visual Motion Analysis.

"Hidden Topic Markov Models",
Amit Gruber, Michal Rosen-Zvi and Yair Weiss,
Artificial Intelligence and Statistics (AISTATS), San Juan, Puerto Rico, March 2007.
Source code abavailable now, See above.
Take a look at the Hidden Topic Markov Model Page.

"Incorporating non-motion cues into 3D motion segmentation",
Amit Gruber and Yair Weiss,
European Conference on Computer Vision (ECCV), Graz, Austria, May 2006.

"Incorporating constraints and prior knowledge into factorization algorithms - an Application to 3D recovery",
Amit Gruber and Yair Weiss,
Subspace, Latent Structure and Feature Selection techniques: Statistical and Optimization perspectives Workshop, Bohinj, Slovenia, February 2005.

"Multibody Factorization with Uncertainty and Missing Data Using the EM Algorithm",
Amit Gruber and Yair Weiss,
Computer Vision and Pattern Recognition (CVPR), Washington D.C., June 2004.

"Factorization with Uncertainty and Missing Data: Exploiting Temporal Coherence",
Amit Gruber and Yair Weiss,
Neural Information Processing Systems (NIPS), Vancouver, December 2003.