[Imageworld] Research Engineer Position at INRIA Willow, Paris,
ivan.laptev at inria.fr
Tue Apr 10 00:47:18 CEST 2012
Research Engineer Position, 1 year (possible extension to 2 years)
Topic: Detection, recognition and tracking of people in video.
INRIA Willow is pursuing research in video analysis. We target realistic
and large-scale video data (TV, movies, YouTube) and focus on the
recognition of people, their actions, person-scene and person-object
interactions. This research has resulted in a number of recent
publications at ICCV11, NIPS11, ICCV09, CVPR09, CVPR08. We have also
proposed Action Classification Competition at PASCAL VOC 2010 and
co-organized workshops and tutorials on human analysis, e.g., ECCV10
Tutorial on Human Action Recognition and CVPR11 Workshop on Gesture
To support and strengthen our research efforts, we are looking for an
engineer to work on the detection, recognition and tracking of people in
realistic video. Based on our prior work CVPR09, the first goal will be
to create a system able to reliably detect and track people in hundreds
of hours of video. The second goal will be to extend this system and to
recognize detected people (face recognition) as well as their traits
(gender, age, etc.). For videos accompanied by scripts, we also aim to
design an automatic annotation system which will learn and recognize
actors from their visual appearance in video and their names in
associated scripts in a fully automatic fashion. The work in this
project will build on and will integrate our recently developed research
code while extending and improving it for addressing the challenge of
video analysis at a large scale.
WILLOW (http://www.di.ens.fr/willow/) is a joint research group between
Institut National de Recherche en Informatique et Automatique (INRIA),
l'École Normale Superieure (ENS) and Centre National de la Recherche
Scientifique (CNRS) - three strong research institutions in France.
Based in central Paris, WILLOW currently includes four permanent
researchers, several post-docs and visitors as well as a number of PhD
and Master students working in different areas of computer vision and
We encourage applications from strong and motivated candidates with
- Master or PhD degree in computer science (Grandes École profile for
- Prior experience in Computer Vision,
- Strong programming skills and experience with C/C++ and Matlab,
- Fluent written and spoken English.
Prior knowledge in the following topics will be considered as an
advantage: large-scale data analysis; video and image analysis; visual
recognition and machine learning. Successful candidates will have an
opportunity to contribute to the state-of-the-art research by publishing
papers at the major venues of computer vision. The salary scale depends
on prior experience and ranges between 2,000-2,500 Euros Net per month.
To apply, please send your CV together with the names of two referees to
Ivan Laptev <ivan.laptev at inria.fr> and Josef Sivic <Josef.Sivic at ens.fr>.
More information about the Imageworld