Parametric Information Maximization for Generalized Category Discovery
Abstract
We introduce a Parametric Information Maximization (PIM) model for the Generalized Category Discovery (GCD) problem. Specifically, we propose a bi-level optimization formulation, which explores a parameterized family of objective functions, each evaluating a weighted mutual information between the features and the latent labels, subject to supervision constraints from the labeled samples. Our formulation mitigates the class-balance bias encoded in standard information maximization approaches, thereby handling effectively both short-tailed and long-tailed data sets. We report extensive experiments and comparisons demonstrating that our PIM model consistently sets new state-of-the-art performances in GCD across six different datasets, more so when dealing with challenging fine-grained problems. Our code: https://github.com/ThalesGroup/pim-generalized-category-discovery.
Cite
Text
Chiaroni et al. "Parametric Information Maximization for Generalized Category Discovery." International Conference on Computer Vision, 2023. doi:10.1109/ICCV51070.2023.00166Markdown
[Chiaroni et al. "Parametric Information Maximization for Generalized Category Discovery." International Conference on Computer Vision, 2023.](https://mlanthology.org/iccv/2023/chiaroni2023iccv-parametric/) doi:10.1109/ICCV51070.2023.00166BibTeX
@inproceedings{chiaroni2023iccv-parametric,
title = {{Parametric Information Maximization for Generalized Category Discovery}},
author = {Chiaroni, Florent and Dolz, Jose and Masud, Ziko Imtiaz and Mitiche, Amar and Ayed, Ismail Ben},
booktitle = {International Conference on Computer Vision},
year = {2023},
pages = {1729-1739},
doi = {10.1109/ICCV51070.2023.00166},
url = {https://mlanthology.org/iccv/2023/chiaroni2023iccv-parametric/}
}