Unsupervised Shadow Removal Using Target Consistency Generative Adversarial Network
This addresses the problem of removing shadows from images without paired data for computer vision applications, representing a strong incremental improvement over existing unsupervised methods.
The paper tackles unsupervised shadow removal by proposing a target-consistency generative adversarial network (TC-GAN) that learns a one-sided mapping from shadow to shadow-free images, outperforming state-of-the-art unsupervised methods by 14.9% in FID and 31.5% in KID and achieving comparable performance to supervised methods.
Unsupervised shadow removal aims to learn a non-linear function to map the original image from shadow domain to non-shadow domain in the absence of paired shadow and non-shadow data. In this paper, we develop a simple yet efficient target-consistency generative adversarial network (TC-GAN) for the shadow removal task in the unsupervised manner. Compared with the bidirectional mapping in cycle-consistency GAN based methods for shadow removal, TC-GAN tries to learn a one-sided mapping to cast shadow images into shadow-free ones. With the proposed target-consistency constraint, the correlations between shadow images and the output shadow-free image are strictly confined. Extensive comparison experiments results show that TC-GAN outperforms the state-of-the-art unsupervised shadow removal methods by 14.9% in terms of FID and 31.5% in terms of KID. It is rather remarkable that TC-GAN achieves comparable performance with supervised shadow removal methods.