Extremely Weak Supervised Image-to-Image Translation for Semantic Segmentation

TL;DR

This paper introduces a method for image-to-image translation that requires very few paired training samples, significantly improving performance over unsupervised methods and approaching fully supervised results with minimal data.

Contribution

It proposes a novel approach to select extremely few paired samples and combine them with unpaired data to enhance I2I translation for semantic segmentation.

Findings

Using only one paired sample outperforms CycleGAN.

Performance approaches pix2pix trained on thousands of pairs.

Combining selected pairs with unpaired data boosts results.

Abstract

Recent advances in generative models and adversarial training have led to a flourishing image-to-image (I2I) translation literature. The current I2I translation approaches require training images from the two domains that are either all paired (supervised) or all unpaired (unsupervised). In practice, obtaining paired training data in sufficient quantities is often very costly and cumbersome. Therefore solutions that employ unpaired data, while less accurate, are largely preferred. In this paper, we aim to bridge the gap between supervised and unsupervised I2I translation, with application to semantic image segmentation. We build upon pix2pix and CycleGAN, state-of-the-art seminal I2I translation techniques. We propose a method to select (very few) paired training samples and achieve significant improvements in both supervised and unsupervised I2I translation settings over random selection. Further, we boost the performance by incorporating both (selected) paired and unpaired samples in the training process. Our experiments show that an extremely weak supervised I2I translation solution using only one paired training sample can achieve a quantitative performance much better than the unsupervised CycleGAN model, and comparable to that of the supervised pix2pix model trained on thousands of pairs.