Education

PhD in Computer Science

Title: Variational principles for tracking color object contours in video

Supervisors : Henri Maître and Jean-Marie Monget.

Date: February 1991

School: Mines Paris Tech

Habilitation (HDR)

Title: Automatic film analysis – from image sequences to action sequences.

Readers: Roger Mohr (Grenoble University), Jean Ponce (Ecole Normale Supérieure) and Andrew Zisserman (Oxford University).

Date : December 2009

School : Grenoble University.

Masters in Engineering

Mines Paris Tech, 1983-1986.

Professional experience

Prizes

Computer Vision and Image Understanding 2013 Most cited paper award for « A survey of vision-based methods for action representation, segmentation and recognition ».

French National Prize for Startup Creation in June 2006 (by international jury of experts with Bill Gates as a president).

2^nd Best Paper Award Eurographics in 1994 : « Triangulating multiply-connected polygons: A simple, yet efficient algorithm ».

Research supervision

Internships at INA:  

• Rotoscopy: Tracking deformable contoursfrom video for cartoon animation (Jean-Philippe Malkasse, Telecom Paris Tech, 1996).
• MPEG2 audio decoder for silence-music-speech  discrimination (Telecom Paris Tech, 1997).
• Shot segmentation in MPEG1 video, with Françoise Préteux (Séverine Pelosse, Telecom Paris Tech, 1998-1999).
• Annotation of INA AIM corpus of broadcast news, with Olivier Chevet (Julien Guerraz, 1999).
• An authoring tool for film templates (Yoann Blanchet, IMAC, 1999-2000).
• Silence-music-speech discrimination using mathematical morphology, with Jean Serra (Discrimination parole/musique/bruit par analyse morphologique et granulométrie de spectres de fréquence (Andres Lafuente Fernandez, Mines Paris Tech, 2000).

Internships at INRIA :

• Actor recognition in movies (Xavier Le Francois , 2002).
• Movie/script segmentation and alignment, with Cécile Roisin (Mickael Viougeas and Cédric Riou, 2002).
• Client-server environment for reading movies, with Cécile Roisin (Mickael Viougeas and Cédric Riou, 2002).
• Automatic classification of camera movements in movies (Caroline Le Corvec, 2003).
• Tracking of actors in movies (Benjamin Chastagnier, 2003).
• Action recognition from dynaic 3D volumes, with Edmond Boyer (Corinne Serose, 2004).
• HMM-based shot segmentation in movies (Mélaine Gautier, 2004).
• Multi-Caméra Monitoring System, with Frederic Devernay (Florian Geffray and Elodie Sannier, 2004).
• Multi-camera gesture recognition (Mélaine Gautier, 2005).
• Calibration of 3-D human skeletal dimensions from images, with Loic Lefort (Yannic Ropars, 2005).
• Wifi-based remote controler for multi-camera monitoring, with Frederic Devernay (Corentin Baron and Émilien Kia, 2006).
• Silhouette interpolation from trinocular camera, with Gabriel Taubin (Doug Lanman, Brown University, 2006).
• Temporal statistics for background subtraction in 3D video, with Chad Jenkins (Matt Loper, Brown University, 2006).
• Strategic analysis of startup project [SEE-NUSH], with Jean-Pierre Boissin (Soraya Khodja, Laetitia Littoz-Monnet, Adrien Chatenoud, IAE Grenoble, 2006).
• Methodology for analyzing sart-up project [SEE-NUSH], with Jean-Pierre Boissin (Ludovic Boisson, IAE Grenoble, 2006).
• Tracking actors with a moving camera (Rémy Lopez and Indy Estublie, 2010).
• A virtual cinematography system that learns from examples, with Marc Christie (Mathieu Chollet, INA Rennes, 2011).
• Statistical shape grammars for 3D animation (Quentin Doussot, 2012).
• 2.5D Rotoscoping — A Space-Time Topology Approach (Boris Dalstein, 2012).
• Automatic shot and scene segmentation in classical Hollywood movies (Vaibhav Singh, 2012).
• Multi-touch choreography, with Marie-Paule Cani et Jarek Rossignac (Mark Luffel, Georgia Tech, 2012).

PhD thesis co-supervised at INA:

• Classification of audiovisual sequences, with François Pachet (Jean Carrive. Paris University, 2000).

PhD theses co-supervised at INRIA :

• Markerless multi-camera motion capture, with Radu Horaud (David Knossow, 2007).
• Action Representation and Recognition, with Edmond Boyer (Daniel Weinland, 2008).

PhD theses in progress :

• Semantically-aware video editing of theatre performances (Vineet Gandhi, 2014).
• Sketching 3D Character Animation, with Marie-Paule Cani (Martin Guay, 2015).
• Automatic cinematography in crowded urban scenes, with Marc Christie (Quentin Galvane, 2015).
• Directing virtual actors by interaction and mutual imitation, with Gerard Bailly (Adela Barbulescu, 2015).
• Camera models and algorithms for creating stereoscopic 3-D video, with Frédéric Devernay (Sergi Pujades Rocamora, 2015).

PhD jurys :

• Segmentation and semantic indexing of video documents, with applications in broadcast news. Claire-Hélène Demarty, Mines Paris Tech, January 2000 (examiner).  
• From text to generation of virtual 3D environments, with applications to theatre scenography. Tahiry Andriamarozakaniaina, University of Toulouse, September 2012 (reader).
• Vision-based tracking of flexible and articulated objects. Daniel Wesierski, Telecom Paris Tech, March 2013 (reader).
• Virtual camera control based on dynamic spatial partitions. Christophe Lino, University of Britanny, October 2013 (examiner).

Responsibilities 

International responsibilities :

• Member of MPEG (Motion Picture Expert Group) in 1997-2000.
• Co-chair of ICCV workshop  on « Modeling People and Human Interaction » in Beijing, China, in 2005 (with Pascal Fua, Adrian Hilton, Daniel Gatica-Perez and Marian Stewart Bartlett).
• Co-chair of 3 international workshops on « 3-D Cinématography » with Gabriel taubin (New York University, June  2006, Banff Center, Canada, July 2008 and Providence, USA, June 2012).  The first workshop was featured in a special issue of IEEE Computer Graphics and Applications in mai 2007. I co-edited a book published by Springer-Verlag (2010) with the proceedings of the second workshop.
• Chair of forth-coming  AAAI workshop on « Intelligent Cinematography and Editing » (WICED), Quebec City, Canada, May 2014.

National responsibilities :

• Member of the working group on « Multimedia indexing » (AIM), CNRS GDR ISIS, in 1998-2000.
• Co-chair of working group on “Action and gesture”, CNRS GDR ISIS, in 2006-2007 and since 2011 (with Catherine Achard and Patrick Horain).
• I served as president of the jury for Master 2 students at the University of Grenoble, Mathématiques et Informatique, option Graphique, Vision et Robotique (GVR) in 2013.

Program committees :

• International Conference on Motion in Games  (MIG).
• International Conference on Interactive Storytelling (ICIDS).
• ACM conference on Computational Aesthetics (CAE).
• Workshop on Intelligent Cinematography and Editing (WICED).
• Workshop on Intelligent Narrative Technologies (INT).
• International Conference on 3D Imaging (IC3D).

I am a regular reviewer for IJCV, IEEE PAMI, CVIU, IEEE Multimedia, ACM SIGGRAPH,  BMVC and CVPR.

Management

Team leadership :

• From 1997 to 2000, I created and managed a team of 4 researchers and engineers in content-based video indexing at INA.
• From 2007 to 2009,  I managed a team  of 5 researchers and engineers in virtual cinematography at Xtranormal Technologies, delivering the Magicam system which was put in production from 2008 to  2013.
• From 2005 to 2007, I managed the international team VAMP (Video and Mesh Processing for 3D Cinematography), a international collaboration between INRIA and  Brown University (Providence,USA).
• Since 2011,  I am the director of the « Image and Geometry » department of Laboratoire Jean Kuntzman, UMR CNRS 5224 , Grenoble University.

Project management :

• Project leader at INA for European project DiVAN (1998-2000): Distributed Video archive Network.  I took part in the writing of the grant proposal, the management of the project , and the development of a prototype system for indexing video collections .
• Project leader at INA for RIAM project SEMOCAP (2004-2006) : Markerless motion capture for independent 3-D animation.
• Project leader for CIBLE  project SCENOPTIQUE (2011-2014) : I wrote the grant proposal and I am leading this a collaboration between INRIA and Theatre des Celestins, Lyon,
• Project leader at INRIA for national « investing in the future » project ACTION 3DS (2011-2014) : I took part in the writing of the grant proposal and I am leading a team of 4 researchers developping novel tools for directing stereoscopic movies..
• Project leader at INRIA for  ANR-funded project  CHROME (2012-2015) : I took part in the writing of the proposal and I am leading a team of  4 researchers developping novel tools for virtual  cinematography  in crowded environments.
• Project leader at INRIA for ANR funded project  SPECTACLENLIGNES (2013-2015) : I took part in the writing of the grant proposal  and I am leading a team of 3 researchers on  audio-visual indexing of theatre rehearsals.
  

International collaborations

I have spent a total of 5 years in the US and Canada :

• Post-doc in 1992-1993 at  IBM T.J Watson Research Center, Yorktown, New York.
• Visiting scientist in 2000-2001 at IBM T.J Watson Research Center, Yorktown, New York.
• Project leader in 2007-2009 at Xtranormal Technologies, Montréal.

Teaching

In addition to my research work in industry and academia, I have spent a total of 400 hours teaching different topics in computer science to diverse audiences :

• Data analysis at Telecom Paris Tech, 1990-91.
• Data analysis at Mine Paris Tech and Imperial College, London, 1990-91.
• Theoretical computer science and algorithmics, ENSAE, 1994-96. I created and wrote this course.
• Content-based video indexing,  at INA-Formation, Ecole des Ponts, and Mines Paris Tech in 1998-2000.
• Human-Computer Interaction (HCI), Grenoble University, 2002-2004.
• Computer Vision, Grenoble University, 2002-2004.
• Gesture recognition, summer school of the University of Fribourg, Switzerland, 2010.
• Advanced 3-D animation , Grenoble University, 2011 and 2012.

Dissemination

While at INA, I presented the scientific challenges of content-based video indexing  to various  audience, including the European Broadcasting Union (EBU), the International Federation of Television Archives (IFTA) and the Collège de Polytechnique.

In 2011, I presented my research team at the Experimenta Art and Sciences Forum in Grenoble.

In 2012, I presented my research program on  « directing virtual worlds » on the INRIA web site  (www.inriality.fr/culture-loisirs/art/virtuel/mettre-en-scene-le/).

In 2013, I organized an INTECH/INRIA workshop on the convergence between digital cinéma and interactive games (http://www.inria.fr/centre/grenoble/actualites/cinema-et-jeu-video-la-convergence-passe-par-inria).