Talks |

Home |

Publications |

Awards |

Teaching |



Hichem Sahbi, PhD, HDR


Current Position
CNRS Researcher and Professor (and HDR)

UPMC, Sorbonne University

Paris, France

Voice: +33 1 44 27 88 18
Email: first name (dot) family name (at) lip6 (dot) fr

Positions and Visits
Currently: CNRS Researcher/Professor (and HDR) at the LIP6 Lab, UPMC, Sorbonne University France.
2007-2017: CNRS Researcher/Professor (and HDR) at the LTCI Lab, TELECOM ParisTech, France.
2006-2007 Senior Research Associate at ENPC (as Chargé de Recherche), Certis Lab, ENPC (joint with ENS-ULM and INRIA-Sophia) , France.
2005-2006 Research associate at the Machine Intelligence Lab, Cambridge University, UK.
2004-2005 Research assistant at INRIA-Rocquencourt, France.
2003-2004 Post doctoral fellow at the Fraunhofer Gesellschaft (GMD Institute), Darmstadt, Germany.
2000-2003 PhD. Candidate with professor Donald Geman , INRIA Rocquencourt, France (defense 07 April 03.)
2001-2002 Invited visiting PhD student at the Center for Imaging Science, Johns Hopkins University, (December 2001 and September 2002)
2009 Invited visiting associate professor at the Institute of Automation, Chinese Academy of Science, Beijing, China (October 2009)
2012 Invited visiting researcher at the "Google Faculty Summit", Google, London, UK (September 2012)


Selected Talks and Lectures

  • Deep learning and visual recognition lecture (video), invited speaker at the Deep Learning Workshop, nov 2018 (slides in english: PDF)
  • Introduction to statistical learning and kernel methods, invited speaker at the Machine Learning Summer School (AMLSS), june 2018 (slides in english: PDF talk I and PDF talk II)
  • Finite state machines for structured scene decoding (video in english), at RLCO - NIPS, dec 2014
  • All Talks

Committees, Memberships, Scientific and Administrative Responsibilities


Research Interest

  • Statistical Machine Learning: manifold learning and graph based inference, kernel methods and kernel design.
  • Deep Learning: deep kernel learning, deep representation and attribute learning.
  • Coarse to Fine Testing: hierarchy of classifiers for object detection in vision.
  • Clustering: fuzzy approach, Gibbs distributions and simulated annealing.
  • Generalization Bounds: model selection and parameter estimation.


Some Selected Publications (Rank A/A+ conferences, ref journals, Best Paper Award publications ...)

  • H. Sahbi and D. Geman. A hierarchy of support vector machines for Face detection. (JMLR PDF link) Journal of Machine Learning Research (JMLR: ref journal in machine learning), Volume 7, 2087--2123, October, 2006.
  • H. Sahbi, J-Y Audibert and R. Keriven. Graph-Cut Transducers for Relevance feedback in Content Based Image Retrieval. (PDF) In the proceedings of the International Conference on Computer Vision (ICCV: rank A/A+ conf in computer vision), October 2007.
  • H. Sahbi. Kernel PCA for Similarity Invariant Shape Recognition. (PDF), (DL) Neurocomputing, Volume/Issue 70/16-18 pages 3034-3045, November 2007.
  • H. Sahbi, J-Y. Audibert and R. Keriven. Context-Dependent Kernels for Object Classification (PDF), (CS Digital Library). In Pattern Analysis and Machine Intelligence (PAMI: ref journal in computer vision and learning ), number 33, volume 4, April 2011.
  • H. Sahbi, J-Y. Audibert, J. Rabarisoa and R. Keriven. Robust Matching and Recognition using Context-Dependent Kernels. (PDF), In the proceedings of International Conference on Machine Learning (ICML: rank A/A+ conf in machine learning), July 2008.
  • H. Sahbi, P. Etyngier, J-Y Audibert and R. Keriven. Manifold Learning using Robust Graph Laplacian for Interactive Image Retrieval. (PDF), In the proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR: rank A/A+ conf in computer vision), June 2008.
  • H. Sahbi, J-Y. Audibert, J. Rabarisoa and R. Keriven. Context-Dependent Kernel Design for Object Matching and Recognition. (PDF), In the proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR: rank A/A+ conf in computer vision), June 2008.
  • H. Sahbi, J-Y Audibert, J. Rabarisoa and R. Keriven. Object Matching and Recognition Using Context Dependent Similarity Kernels (Best Regular Paper Award by Queen's Marry University of London), In the proceedings of IEEE Workshop on Content Based Multimedia Indexing (CBMI), June 2008.
  • H. Sahbi and X. Li. Context Based Support Vector Machines for Interconnected Image Annotation (oral). (PDF). ("The Saburo Tsuji" Best Regular Paper Award). In the Asian Conference on Computer Vision (ACCV), 2010. (oral paper acceptance rate, 4.7%)
  • H. Sahbi and X. Li. Context Dependent SVMs for Interconnected Image Network Annotation. (PDF) In the proceedings of (ACM MULTIMEDIA: rank A/A+ conf in multimedia and image search ), October 2010.
  • L. Wang and H. Sahbi. Directed Acyclic Graph Kernels for Action Recognition. (PDF) In the proceedings of the International Conference on Computer Vision (ICCV: rank A/A+ conf in computer vision), December 2013.
  • P. Vo and H. Sahbi. Spacious: an interactive mental search interface. (DL) The 36th International ACM SIGIR conference on research and development in Information Retrieval, (SIGIR: rank A/A+ conf in information retrieval ), July 2013.
  • P. Vo and H. Sahbi. Semantic Subspace Learning for Mental Search in Satellite Images (PDF) (Best Paper Award Nomination), The International Geoscience and remote sensing symposium, IGARSS, July 2013.
  • H. Sahbi, L. Ballan, G. Serra and A. Del Bimbo. Context-Dependent Logo Matching and Recognition (PDF) In IEEE Transactions on Image Processing, volume 22, number 3, pages 1018-1031, 2013.
  • H. Sahbi. Finite State Machines for Structured Scene Decoding (PDF) In RLCO, Neural Information Processing Systems (NIPS: rank A/A+ conf), 2014.
  • H. Sahbi. ImageCLEF annotation with explicit context-aware kernel maps (PDF) In International Journal of Multimedia Information Retrieval (IJMIR, Springer), 2015.
  • M. Jiu and H. Sahbi. Nonlinear Deep Kernel Learning for Image Annotation, (PDF) In the IEEE Transactions on Image Processing (IEEE TIP, volume 26, issue 4), april 2017.
  • H. Sahbi. Interactive Satellite Image Change Detection with Context-Aware Canonical Correlation Analysis, (PDF) the IEEE Geoscience and Remote Sensing (IEEE GRSL, volume 14, issue 5), may 2017.
  • A. Dutta and H. Sahbi. Stochastic Graphlet Embedding, (PDF) IEEE Transactions on Neural Networks and Learning Systems (accepted, to appear, 2019) (IEEE TNNLS), 2019.
  • M. Jiu and H. Sahbi. Deep Representation Design from Deep Kernel Networks, (PDF) Pattern Recognition (PR, volume 88), april 2019.
  • All Publications
  • Google scholar

SOFTWARES and Databases

Past and Current Projects

Supervised PhD students (as thesis director)

  • 2010-2012: Fei Yuan (PhD Student at TELECOM ParisTech in collaboration with NLPR, The Chinese Academy of Science, currently researcher at NLPR ): Graph based techniques for Event Detection and Recognition (thesis defense june 2012)
  • 2010-2013: Nicolas Bourdis (CIFRE PhD Student at TELECOM ParisTech and EADS, currently researcher at AIRBUS): change detection, (thesis defense, may 2013).
  • 2015-2018: Quentin Oliveau (CIFRE PhD Student at TELECOM ParisTech, currently researcher at DCNS Group): Attribute Learning for Object Recognition (thesis defense, april 2018)
  • 2016-2020: Tristan Postadjian (PhD Student between IGN Paris and UPMC SU, currently researcher at Sodexo Group): High Resolution Remote Sensing Satellite Imagery for Soil Occupancy (thesis defense, feb 2020)
  • 2017-now: Ahmed Mazari (PhD Student at LIP6 lab, Sorbonne University): Deep learning for action recognition (ongoing)
  • 2019-now: Haoming Zhan (PhD Student at LIP6 lab, Sorbonne University): Deep continual learning (ongoing)

Supervised post docs (as official postdoc director)

  • Oct 2014-Oct 2015: Anjan Dutta (former Post doc at LTCI, Telecom ParisTech, Paris): graph hashing for pattern classification

Supervised and Visiting Students (as Co-Supervisor)

  • 2003-2007: Liu Huajian (former PhD candidate at Fraunhofer IPSI, defense 2007, currently Research Staff at Fraunhofer SIT, Germany ): "object of interest" based image authentication
  • 2004-2007: Stefan Thiemert (Former PhD Candidate at Fraunhofer IPSI, Germany): content-based video watermarking
  • 2004: Varun Kumta (M.S. Candidate IIT INDIA currently at IBM INDIA): Robustness of image retrieval to image watermarking
  • 2007: Bhavna Agarwal (M.S. Candidate from INDIA): Image Matching and Recognition
  • 2008: Olivier Le Floch (B.S Candidate from l'Ecole Polytechnique, Palaiseau): Image Search
  • 2007-2008: Jiayu Chen (Former Visiting PhD Student at TELECOM ParisTech from University of Wuhan): Image Retrieval
  • 2005-2006: Khe Chai Sim (Former Post Doc at Cambridge university moved to A-Star and currently professor at Institute of Digital Media, Singapore ): System combination for statistical machine translation
  • 2006-2007: Patrick Etyngier (Former PhD. student at Certis, ENPC, France, defense Jan 2008, moved to Philips Paris): Graph Laplcian for Interactive Image Retrieval
  • April-Aug 2012: Surrender Varma (Former PhD vistor during 5 months from IIT INDIA): change detection
  • 2011-2012: Malek Nadil (Former PhD visitor during 10 months from USTHB Algiers): face processing
  • 2012-2014: Emilie Au (PhD Student at LIP6 and TELECOM ParisTech): Learning on graphs


  • Dr. Hichem Sahbi received his M.S. degree in theoretical computer science from the University of Paris Sud, in Orsay, France in 1999, and his PhD in computer vision and machine learning from INRIA/Versailles University, France in 2003. From 2003-2006 he was a research associate first at the Fraunhofer Institute in Darmstadt, Germany, and then at the Machine Intelligence Laboratory at Cambridge University, UK. From 2006-2007, he was a senior research associate at l'Ecole des Ponts Paris Tech, in Paris. Since 2007, he has been a CNRS professor (and HDR) at Telecom ParisTech and then at UPMC, Sorbonne University, in Paris. Dr. Sahbi has participated in various scientific committees, industrial, national, and international projects, and he is the author of many publications. Dr. Sahbi's research interests include statistical machine and deep learning, computer vision, and image retrieval.





    © 2018 LIP6 Lab, Sorbonne University, Paris, France.
    All rights reserved.