WILLOW is based in the
Laboratoire d'Informatique de l'École Normale Superiéure (CNRS/ENS/INRIA UMR 8548) and is a joint research team between
INRIA Rocquencourt,
École Normale Supérieure
de Paris and
Centre National de la Recherche Scientifique.
Our research is concerned with representational issues in visual object
recognition and scene understanding. Our objective is to develop geometric,
physical, and statistical models for all components of the image interpretation
process, including illumination, materials, objects, scenes, and human
activities. These models will be used to tackle fundamental scientific
challenges such as three-dimensional (3D) object and scene modeling, analysis,
and retrieval; human activity capture and classification; and category-level
object and scene recognition. They will also support applications with high
scientific, societal, and/or economic impact in domains such as quantitative
image analysis in domains such as archaeology and cultural heritage
conservation; film post-production and special effects; and video annotation,
interpretation, and retrieval.
Moreover, machine learning now represents a significant part of computer vision
research, and one of the aims of the project is to foster the joint development
of contributions to machine learning and computer vision, together with
algorithmic and theoretical work on generic statistical machine learning.
Research themes
We follow four main research directions:
- 3D object and scene modeling, analysis, and retrieval:
This part of our research addresses the problem of acquiring high-accuracy
geometric models of complex 3D objects from multiple images, and using
these models in rigid and nonrigid registration and retrieval tasks.
- Human activity capture and classification:
This includes combining object detection and tracking to associate
multiple ocurrences of faces and bodies in video footage, and constructing
and recognizing spatio-temporal models of human activities and
interactions.
- Category-level object and scene recognition:
This part of our research is concerned with developing geometric and
statistical models of objects, scenes, and their components that support
semi-supervised learning, effectively handle image variability, and
efficiently support inference.
- Machine learning:
The team focuses on both algorithmic aspects related to vision
applications and theoretical aspects: kernel methods, sparse methods,
calibration methods, graphical models and multi-task learning.