UnLIR: Unsupervised Learning of Image Representation is a JCJC ANR project coordinated by Ronan Sicre starting in January 2020 and ending in june 2024.

The project deals with latent representation learning for complex recognition tasks, such as fine-grained recognition; Unsupervised representation learning; interpretability of deep neural Networks.

The Team is composed of Thierry Artières (LIS), Stephane Ayache (LIS), Yannis Avrithis (INRIA Rennes) and Ronan Sicre (LIS).