PamiTC Job Board - Posting Details

Title: Research Engineer Position - Dynamic Scene Reconstruction, Virtual / Augmented Reality - MPI for InformaticsPosted: May 9, 2016
Company/Institution: Max-Planck-Insitute (MPI) for Informatics
Location: Saarbruecken, Germany
Department: Graphics, Vision & Video

Description:

MPI for Informatics: Research Engineer - Marker-less Motion and Performance Capture, Virtual and Augmented Reality, 3D Video

The research group Graphics, Vision & Video (GVV) (http://gvv.mpi-inf.mpg.de) and the Computer Graphics group at the Max-Planck-Institute for Informatics in Germany are looking for a Research Engineer. MPI for Informatics is one of the world's leading research institutes Computer Science. Visual Computing is a strong research focus and several research groups are contributing to the forefront of Computer Graphics and Computer Vision.

In one of our main lines of research, we are developing groundbreaking new methods for marker-less dynamic scene reconstruction (geometry, motion and appearance) from camera input. This includes some of the first methods for high-quality free-viewpoint video of humans, high-quality performance capture methods for humans in general clothing, detailed dynamic face reconstruction from video, and real-time marker-less motion capture of full-body and hand motion.

See: http://gvv.mpi-inf.mpg.de for recent projects.

The award-winning startup company The Captury is a spin-off of the GVV group at MPI for Informatics and builds up on fundamental research results from the institute. More recently, we started to investigate new methods to capture detailed models of dynamic scenes, including models of reflectance and illumination in the scene, in real-time. Examples of our recent results are the Face2Face real-time face performance capture and reenactment approach which was reported on widely in the media.

Other recent examples from our work are one of the state-of-the-art methods for full articulated hand tracking from a single depth camera, as well as one of the first methods for live intrinsic video.

Handtracker
Live Intrinsic Video

Our mission is to find foundational algorithmic answers to challenging open questions in dynamic reconstruction of general scenes, and publish our results in the top conferences and journals in computer graphics and computer vision. This way, we lay the methodical foundations for interactive perceiving visual computing systems of the future. It is our goal to think beyond what is currently possible and what will be state-of-the-art interactive visual computing systems in many years from now. Our research is thus not geared towards specific applications, but our results will be beneficial in many areas of application. On the one hand, this is content creation for VFX in movies, games and virtual environments. However, our algorithms also benefit future developments in marker-less motion capture, human-computer and human-robot interaction, autonomous systems, biomechanics and sports science, virtual and augmented reality, as well a 3D and free-view point video, e.g. for virtual replay etc.

To develop and validate new methods for dynamic scene reconstruction, we depend on a variety of sophisticated software and hardware systems, as well as specialized measurement setups, such as motion capture systems, shape scanners, and multi-view camera systems, virtual and augmented reality setups. Effective operation of these systems is in many cases a key requirement for research progress. The research engineer will be - on the one hand - responsible for operating, installing and modifying the above hardware systems. He/she will support the team of scientists to achieve their research goals and perform experiments. The engineer will also be actively involved in developing new technology and experimental software prototypes.

A successful applicant should have the following key qualifications:

  • A Bachelor or Master degree or equivalent in computer science, computer animation, media production, or electrical engineering.
  • A background in computer graphics and computer vision techniques and algorithms.
  • Programming experience in C++ (and standard graphics APIs, e.g. OpenGL), and scripting languages, e.g. Python etc.
  • Experience with standard animation tools, such as Maya and 3DS Max, ideally also with the respective APIs.
  • Great problem solving skills.
  • The ability to work independently on multiple tasks simultaneously.
  • Fluency in written and spoken English.
  • The willingness to participate in discussions with researchers and to support them in developing new ideas.

In addition to the requirements above, some of the following qualifications would be a plus:

  • Experience in the development of simple electronic circuits.
  • Experience with a graphics API like OpenGL, with GPU programming e.g. CUDA, as well as with low level control of hardware.
  • Experience with operating and maintaining camera systems in professional production scenarios.
  • Experience in using and operating an optical motion capture system/multi-view video system.
  • Experience in working with motion capture data.

MPI offers a very unique international, collaborative research environment in which a successful applicant will be able to contribute his/her skills in a variety of different projects. The position is limited to 2 years, but with the potential of prolongation after the initial period.


Application Instructions:

Please send your complete application material, including a CV, copies of transcripts and certificates, contacts of two references, and a description of previous projects that match the above requirements to

Prof. Dr. Christian Theobalt
Head of Research Group Graphics, Vision & Video
MPI fuer Informatik - Campus E1.4
66123 Saarbruecken
Germany

Email: theobalt@mpi-inf.mpg.de

Deadline for applications is: June 5, 2016