SS-CPGAN: Self-Supervised Cut-and-Pasting Generative Adversarial Network for Object Segmentation

TL;DR

This paper introduces SS-CPGAN, a self-supervised GAN that improves object segmentation and realistic image composition without manual labels by leveraging a U-Net discriminator and pseudo labels.

Contribution

It presents a novel self-supervised approach combined with a U-Net discriminator to enhance foreground object segmentation in GANs without manual annotations.

Findings

Outperforms state-of-the-art methods on benchmark datasets

Learns semantic and structural information through pseudo labels

Generates meaningful masks for object segmentation

Abstract

This paper proposes a novel self-supervised based Cut-and-Paste GAN to perform foreground object segmentation and generate realistic composite images without manual annotations. We accomplish this goal by a simple yet effective self-supervised approach coupled with the U-Net based discriminator. The proposed method extends the ability of the standard discriminators to learn not only the global data representations via classification (real/fake) but also learn semantic and structural information through pseudo labels created using the self-supervised task. The proposed method empowers the generator to create meaningful masks by forcing it to learn informative per-pixel as well as global image feedback from the discriminator. Our experiments demonstrate that our proposed method significantly outperforms the state-of-the-art methods on the standard benchmark datasets.