Generating Images Part by Part with Composite Generative Adversarial Networks
This work addresses image generation for AI and deep learning by introducing a method to decompose image generation into parts, though it appears incremental as it builds on existing GAN frameworks.
The paper tackles the problem of generating complex images by proposing a composite generative adversarial network (GAN) with multiple generators, each producing parts of an image combined via alpha blending, and demonstrates its ability to learn structure unsupervised, such as generating background and face sequentially on a face dataset.
Image generation remains a fundamental problem in artificial intelligence in general and deep learning in specific. The generative adversarial network (GAN) was successful in generating high quality samples of natural images. We propose a model called composite generative adversarial network, that reveals the complex structure of images with multiple generators in which each generator generates some part of the image. Those parts are combined by alpha blending process to create a new single image. It can generate, for example, background and face sequentially with two generators, after training on face dataset. Training was done in an unsupervised way without any labels about what each generator should generate. We found possibilities of learning the structure by using this generative model empirically.