FBC-GAN: Diverse and Flexible Image Synthesis via Foreground-Background Composition
This addresses a bottleneck in image generation for applications requiring varied and customizable outputs, though it is an incremental improvement over existing GAN methods.
The paper tackles the problem of limited diversity in GAN-generated images due to excessive content correlation between foreground and background, and presents FBC-GAN, which generates foregrounds and backgrounds independently and composes them to achieve superior diversity and flexibility in image synthesis.
Generative Adversarial Networks (GANs) have become the de-facto standard in image synthesis. However, without considering the foreground-background decomposition, existing GANs tend to capture excessive content correlation between foreground and background, thus constraining the diversity in image generation. This paper presents a novel Foreground-Background Composition GAN (FBC-GAN) that performs image generation by generating foreground objects and background scenes concurrently and independently, followed by composing them with style and geometrical consistency. With this explicit design, FBC-GAN can generate images with foregrounds and backgrounds that are mutually independent in contents, thus lifting the undesirably learned content correlation constraint and achieving superior diversity. It also provides excellent flexibility by allowing the same foreground object with different background scenes, the same background scene with varying foreground objects, or the same foreground object and background scene with different object positions, sizes and poses. It can compose foreground objects and background scenes sampled from different datasets as well. Extensive experiments over multiple datasets show that FBC-GAN achieves competitive visual realism and superior diversity as compared with state-of-the-art methods.