• The IMAGINE group offers the six following immediate openings (4 PostDoc, 3 PhD)
  
  - PhD position in Computer Science and Computer Vision: Building modeling and image analysis.
  This PhD position in is funded by the CSTB (Scientific and Technical Centre for Building). The goals are: (1) to develop a framework and methodology for developing *parameterized* models of buildings; (2) to apply this methodology to the development of a few actual models concerning features that impact energy saving issues, such as building geometry and openings; (3) to develop prior-based computer vision techniques relying on these "architectural grammars" for the determination of model parameters from pictures. This work requires good general skills in computer science for domain analysis and data modeling issues, as well as specific skills in computer vision. Applicants must also have good communication skills and be fluent in French as part of the work involves interacting with French professionals (building experts and architects) to capture their expertise.
  To apply, please email CV, references and a brief description of research interests to Renaud Marlet
  - Post-Doctoral position in Stereo reconstruction. Semantized 3D urban modelling.
  Imagine and EADS Foundation offer a position in semantized urban modelling. The candidate will study the contribution of grammar-based methods to state-if-the art multiview stereo reconstruction.
  To apply, please email CV, references and a brief description of research interests to Renaud Keriven
  - PhD position in Stereo reconstruction. Film industry.
  Imagine and the post-production company Mikros Image offer an industry-supported PhD position (a.k.a. CIFRE position) in applications of multiview stereo reconstruction for film industry. MSc cancidates presently in France have the opportunity to do their internship with IMAGINE.
  To apply, please email CV, references and a brief description of research interests to Renaud Keriven
  - Post-Doctoral position in Stereo reconstruction. Street 3D modelling.
  Imagine and IGN (French institute for Geographic Institute) offer a position in 3D street modelling from aerial and terrestrial acquisition. The candidate will benefit from the unique acquisition devices of both institutes to model highly accurate street models.
  To apply, please email CV, references and a brief description of research interests to Renaud Keriven
  - Post-Doctoral position in Stereo reconstruction. Accurate reconstruction.
  Innovative multi-view stereo-vision methods will be developped, with a strong acccent on 3-D precision. The research will be conducted in close cooperation with MAP5, the mathematics department of University Paris-Descartes. Applicants should have a PhD in computer science, applied mathematics or electrical engineering. The successful candidate will join a team of leading computer vision researchers in a 3-year project named Callisto funded by ANR, the French funding agency for research. This is a 1 year position.
  To apply, please email CV, references and a brief description of research interests to Pascal Monasse
  - PhD position in Stereo reconstruction. External calibration.
  The aim is to develop innovative algorithms to perform external calibration in multi-view stereovision, that is recover the positions and orientations of all cameras simultaneously. This is one of the early steps of a multi-view stereo pipeline, whose accuracy is crucial for the recovered 3D geometry. Candidates must have an MSc degree in mathematics, computer science or electrical engineering. A strong mathematical or computer science background and interest in computer vision or image processing are required. MSc cancidates presently in France have the opportunity to do their internship with IMAGINE. As a PhD student you will join a leading research team specialized in stereo 3D reconstruction, machine learning and computer science, 20 minutes away from Paris center. This is a full time position of 3 years with possibility to teach computer science at Ecole des Ponts ParisTech. This position is in the context of a project named Callisto and funded by Agence Nationale de la Recherche (ANR), the French funding agency for research. This project involves a consortium of leading researchers in or around Paris in computer vision.
  To apply, please email CV, references and a brief description of research interests to Pascal Monasse
  - Post-Doctoral position in Statistical Learning and Vision. Exploration-exploitation trade-off.
  The position will be funded by a grant from the French National Research Agency (ANR). The goal of the project funded by ANR is to develop methods making the best possible use of available resources, where the term resource can refer to a piece of computational effort (for example CPU time, memory) devoted to the realization of some computation or some real-world effort that has a real, e.g. financial, cost. A building block of the methods are algorithms which perform a good job in balancing resources between exploitation (making the best decision based on our current, but possibly imperfect, knowledge) and exploration (decisions that may appear sub-optimal but which may yield additional information about the unknown parameters, and, as a result, could improve the relevance of future decisions). The goal of this position is two-fold: first, numerous problems linked to the exploration-exploitation trade-off remain open (even in the well-known multi-armed bandit). We want to address them, i.e. propose algorithms solving them and study their theoretical properties (for instance, evaluate the critical amount of resources required to perform the task decently). This part will be mainly theoretical and in the same spirit as the nice book: Prediction, learning, and games by Nicolò Cesa-Bianchi and Gábor Lugosi. Second, we want to address some of the energy minimization problems that arise in Computer Vision, in particular combinatorial problems of 3D reconstruction, by using the proposed algorithms. This work will therefore require strong theoretical skills in probability and statistics, and ability to implement algorithms working on real-world data.
  To apply, please email CV, references and a brief description of research interests to Jean-Yves Audibert