PCGAN: Partition-Controlled Human Image Generation
This work addresses the challenge of generating high-quality human images with controlled backgrounds for applications in computer vision, though it appears incremental by building on existing pose-conditioned methods.
The paper tackles the problem of generating human images with specific poses and backgrounds, which often results in blurred backgrounds in existing methods. The proposed Partition-Controlled GAN generates realistic human images with desired poses and backgrounds, as validated on Market-1501 and DeepFashion datasets.
Human image generation is a very challenging task since it is affected by many factors. Many human image generation methods focus on generating human images conditioned on a given pose, while the generated backgrounds are often blurred.In this paper,we propose a novel Partition-Controlled GAN to generate human images according to target pose and background. Firstly, human poses in the given images are extracted, and foreground/background are partitioned for further use. Secondly, we extract and fuse appearance features, pose features and background features to generate the desired images. Experiments on Market-1501 and DeepFashion datasets show that our model not only generates realistic human images but also produce the human pose and background as we want. Extensive experiments on COCO and LIP datasets indicate the potential of our method.