Semi-Supervised Training of Structured Output Neural Networks with an Adversarial Loss

Mateusz Kozinski (1) , Loïc Simon (1) , Frédéric Jurie (1)

1 Equipe Image - Laboratoire GREYC - UMR6072

We propose a method for semi-supervised training of structured-output neural networks.Inspired by the framework of Generative Adversarial Networks (GAN), we train a discriminator network to capture the notion of `quality' of network output.To this end, we leverage the qualitative difference between outputs obtained on the labelled training data and unannotated data.We then use the discriminator as a source of error signal for unlabelled data.This effectively boosts the performance of a network on a held out test set.Initial experiments in image segmentation demonstrate that the proposed framework enables labelling two times less data than in a fully supervised scenario, while achieving the same network performance.

Mots clés: semi; supervised learning; segmentation; adversarial learning

Domaines: Vision par ordinateur et reconnaissance de formes [cs.CV]

Mateusz Kozinski, Loïc Simon, Frédéric Jurie. Semi-Supervised Training of Structured Output Neural Networks with an Adversarial Loss. ORASIS 2017, GREYC, Jun 2017, Colleville-sur-Mer, France. ⟨hal-01866633⟩