Counter-examples generation from a positive unlabeled image dataset

Florent Chiaroni

doi:10.1016/j.patcog.2020.107527

Counter-examples generation from a positive unlabeled image dataset

2020-11-01

9

OA

AI

F

Florent Chiaroni *

G

Ghazaleh Khodabandelou

DOI：10.1016/j.patcog.2020.107527

Original

Save

View PDF

Abstract

En 中文

This paper considers the problem of positive unlabeled (PU) learning. In this context, we propose a two-stage GAN-based model. More specifically, the main contribution is to incorporate a biased PU risk within the standard GAN discriminator loss function. In this manner, the discriminator is constrained to steer the generator to converge towards the unlabeled samples distribution while diverging from the positive samples distribution. Consequently, the proposed model, referred to as D-GAN, exclusively learns the counter-examples distribution without prior knowledge. Experimental results on simple and complex image datasets demonstrate that our approach outperforms state-of-the-art PU methods without prior by overcoming issues such as sensitivity to prior knowledge or first-stage overfitting. (C) 2020 Elsevier Ltd. All rights reserved.

Keywords:

Generative adversarial networks (GANs)

Generative models

Semi-supervised learning

Partially supervised learning

Deep learning

Counter-examples generation from a positive unlabeled image dataset

Abstract

AI Summary

Journal

Organization