aharon.barhillel@gmail.com

Research scientist
Vision and smart sensing lab
Advanced Technical Center Israel (ATCI)  
General Motors
phone:  Work 972-9-9720617                             

Home: 972-3-5350952        

    
        

I finished my Ph.D in the Interdisciplinary center for neural computation, the Hebrew university of Jerusalem, 2006. My thesis included topics from machine learning and computer vision. It was guided by Prof. Daphna Weinshall . Since then I've been working as a research scientist and a hardware architect in Intel, and now as a research scientist at GM. Research Interests: Machine Learning: Distance function learning, Feature selection and synthesis, Task transfer Computer Vision: Object class recognition and detection Machine vision  applications: Grid management, Spike sorting Publications:   2011 Fusing visual and range imaging for object class recognition  (The Data used in the paper) Aharon Bar-Hillel, Dmitry Hanukaev, Dan Levi ICCV 2011 Recent progress in road and lane detection - a survey Aharon Bar-Hillel, Ronen Lerner, Dan Levi, Guy Raz To appear in Machine Vision and Applications 2011 2010 Part-based feature synthesis for human detection (PFS code is available below) Aharon Bar-Hillel, Dan Levi, Eyal Krupka, Chen Goldberg  ECCV 2010   2009 Efficient human computation: the distributed labeling problem Ran Gilad-Bachrach, Aharon Bar-Hillel, Liat Ein-Dor  Workshop on human computation, KDD 2009  2008 Efficient learning of relational object class models Aharon Bar-Hillel, Daphna Weinshall International Journal of Computer Vision (IJCV), 2008 Gaussian mixture models with equivalent constraints Noam Shental, Aharon Bar-Hillel, Tomer Hertz        , Daphna Weinshall A book chpater in "Constrained Clustering: Advances in Algorithms, Theory and Applications" by Sugatu Basu, Iam Davidson and Kiri Wagstaff (Eds.), 2008 Comparison Processes in Category learning: From Theory to Behavior Rubi Hammer, Aharon Bar-Hillel, Tomer Hertz        , Daphna Weinshall and Shaul Hochstein  Brain Research, Special issue on 'Brain and Vision', 2008 2007 Reinforcement learning for capacity tuning of multi core servers Liat Ein-Dor, Yossi Ittach, Aharon Bar-Hillel, Amir Di-Nur, Ran Gilad-Bachrach In "machine learning for systems problems", workshop in NIPS 2007 Learning distance functions by coding similarity Aharon Bar-Hillel, Daphna Weinshall In the 24rd International Conference on Machine Learning (ICML), 2007. Workstation capacity tuning using reinforcement learning Aharon Bar-Hillel, Amir Di-Nur, Liat Ein-Dor, Ran Gilad-Bachrach, Yossi Ittach In International conference for high performance computing, networking, storage and analysis (SC), 2007  2006 Subordinate class recognition using relational object models Aharon Bar-Hillel, Daphna Weinshall In 20th International Conference on Neural Information Processing Systems (NIPS), 2006. Learning from weak represnentation using distance function and generative models Aharon Bar-Hillel Ph.D Thesis, 2007 Learning a kernel function for classification with small training samples Tomer Hertz        , Aharon Bar Hillel, Daphna Weinshall In the 23rd International Conference on Machine Learning (ICML), 2006 Spike sorting: Bayesian clustering of non stationary data Aharon Bar Hillel, Adam Spiro, Eran Stark         Journal of Neuroscience Methods 157(2), 303-316, October 2006 2005 Efficient learning of relational object class models Aharon Bar Hillel, Tomer Hertz         and Daphna Weinshall International Conference on Computer Vision (ICCV), 2005. Learning a Mahalanobis metric from equivalence constraints Aharon Bar Hillel, Tomer Hertz        , Noam Shental and Daphna Weinshall In Journal of Machine Learning Research 6(Jun): 937-965, 2005 Object class recognition by boosting a part based model Aharon Bar Hillel, Tomer Hertz         and Daphna Weinshall In In IEEE conference on Computer Vision and Pattern Recognition (CVPR), 2005 2004 Spike sorting: Bayesian Clustering of non-stationary data   Aharon Bar-Hillel, Adam Spiro and Eran Stark           In 18th International Conference on Neural Information Processing Systems - NIPS, Vancouver, Canada, 2004. Boosting Margin Based Distance Functions for Clustering   Tomer Hertz        , Aharon Bar-Hillel and Daphna Weinshall  In 21st International Conference on Machine Learning (ICML 2004), Banff, Alberta, Canada July 2004.  Learning Distance Functions for Image Retrieval   Tomer Hertz        , Aharon Bar-Hillel and Daphna Weinshall  In IEEE Conf. on Computer Vision & Pattern Recognition, Washington, DC, June 2004  Computing Gaussian Mixture Models with EM using Equivalence Constraints Noam Shental, Aharon Bar-Hillel, Tomer Hertz         and Daphna Weinshall In 17th International Conference on Neural Information Processing Systems - NIPS, Vancouver, Canada, December 2004. 2003 Learning with Equivalence Constraints, and the relation to Multiclass Classification   Aharon Bar-hillel , Daphna Weinshall  In The Sixteenth Annual Conference on Learning Theory (COLT) , 2003 Learning distance functions using Equivalence Relations   Aharon Bar-hillel , Tomer Hertz        , Noam Shental, Daphna Weinshall  In The Twentieth International Conference on Machine Learning (ICML) 2003 Computing mixture models with EM using side information (Equivalence Constraints)   Noam Shental, Aharon Bar-hillel , Tomer Hertz        , Daphna Weinshall  In "The Continuum from Labeled to Unlabeled Data in Machine Learning and Data Mining", Workshop in ICML 2003 Enhancing image and video retrieval : Learning via Equivalence Constraints   Tomer Hertz        , Noam Shental, Aharon Bar-hillel ,Daphna Weinshall  In proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2003 My Colleagues: Amir Di-Nur  Liat Ein-Dor  Ran Gilad-Bachrach Chen Goldberg  Tomer Hertz         Yossi Ittach Eyal Krupka Dan Levi Noam Shental Adam Spiro Eran Stark         Daphna Weinshall Matlab code: PFS - Predictive Feature Selection for SVM Matlab code Relational part based model for object class recognition - Matlab code A Coding Similarity Matlab implementation. The DistBoost algorithm Spike sorting code The Constrained EM algorithm - This code requires the BNT package to be on the Matlab path. The Constrained EM algorithm - This code is self sufficient (it includes the BNT package) A RCA Matlab implementation. I like: Terry Pratchet Shlomo Gronich Civilization My Love: Naomi Avigail Hagar