A Generative Model of People in Clothing
This enables data-driven generation of clothed people for applications like fashion or virtual environments, representing a novel approach but with incremental technical advancements.
The authors tackled the problem of generating full-body images of people in clothing without relying on 3D scans or complex rendering, by developing a generative model that splits the process into semantic segmentation and conditional image generation, producing realistic samples conditioned on pose, shape, or color.
We present the first image-based generative model of people in clothing for the full body. We sidestep the commonly used complex graphics rendering pipeline and the need for high-quality 3D scans of dressed people. Instead, we learn generative models from a large image database. The main challenge is to cope with the high variance in human pose, shape and appearance. For this reason, pure image-based approaches have not been considered so far. We show that this challenge can be overcome by splitting the generating process in two parts. First, we learn to generate a semantic segmentation of the body and clothing. Second, we learn a conditional model on the resulting segments that creates realistic images. The full model is differentiable and can be conditioned on pose, shape or color. The result are samples of people in different clothing items and styles. The proposed model can generate entirely new people with realistic clothing. In several experiments we present encouraging results that suggest an entirely data-driven approach to people generation is possible.