Description: Invariance and supervision in visual learning
The goal of this PhD is to revisit invariant representations and reconsider the role of supervision in visual learning. In particular, it will investigate the use of existing knowledge from computer vision tasks including matching, retrieval, detection and tracking within more recent deep architectures for visual recognition, with the objective of minimizing the need for supervision in the form of labeled samples.
The candidate should ideally have a degree in Computer Science, Applied Mathematics or Electrical Engineering; solid mathematical background and programming skills; fluency in English language; preferably, prior experience in computer vision, machine learning or data mining.
This call is part of Inria's PhD positions program. The positions are highly competitive: only a limited number will be funded depending on the quality of applications.