Adversarial Learning for Semi-Supervised Semantic Segmentation

TL;DR

This paper introduces an adversarial learning approach for semi-supervised semantic segmentation, utilizing a fully convolutional discriminator to improve accuracy and leverage unlabeled data effectively.

Contribution

It proposes a novel fully convolutional discriminator for semi-supervised segmentation, enabling the use of unlabeled images to enhance model performance.

Findings

Improved segmentation accuracy on PASCAL VOC 2012 and Cityscapes datasets.

Effective semi-supervised learning by identifying trustworthy regions in unlabeled images.

Demonstrated superiority over existing weakly-labeled methods.

Abstract

We propose a method for semi-supervised semantic segmentation using an adversarial network. While most existing discriminators are trained to classify input images as real or fake on the image level, we design a discriminator in a fully convolutional manner to differentiate the predicted probability maps from the ground truth segmentation distribution with the consideration of the spatial resolution. We show that the proposed discriminator can be used to improve semantic segmentation accuracy by coupling the adversarial loss with the standard cross entropy loss of the proposed model. In addition, the fully convolutional discriminator enables semi-supervised learning through discovering the trustworthy regions in predicted results of unlabeled images, thereby providing additional supervisory signals. In contrast to existing methods that utilize weakly-labeled images, our method leverages unlabeled images to enhance the segmentation model. Experimental results on the PASCAL VOC 2012 and Cityscapes datasets demonstrate the effectiveness of the proposed algorithm.