Published in Arxiv Preprint, 2022
In this paper we propose a novel model that addresses the Few-Shot Open-Set Recognition problem with a One-Shot model that is augmented with a discriminator that rejects unknown actions.
Published in Arxiv Preprint, 2022
In this paper, we propose an object shape completion method based on an implicit 3D representation providing a confidence value for each reconstructed point. As a second contribution, we introduce a gradient-based method for efficiently sampling such implicit function at an arbitrary resolution, tunable at inference time.
Published in International Workshop on Continual Semi-Supervised Learning. Springer, Cham, 2022., 2021
This work introduces Distilled Replay, a novel replay strategy for Continual Learning which is able to mitigate forgetting by keeping a very small buffer (1 pattern per class) of highly informative samples.