XingGAN for Person Image Generation
This work addresses person image generation for pose translation, which is incremental as it builds on existing GAN methods with specific architectural improvements.
The authors tackled the problem of generating realistic person images with new poses by proposing XingGAN, a novel GAN that uses two branches for appearance and shape modeling with crossing embeddings, achieving state-of-the-art performance on Market-1501 and DeepFashion datasets.
We propose a novel Generative Adversarial Network (XingGAN or CrossingGAN) for person image generation tasks, i.e., translating the pose of a given person to a desired one. The proposed Xing generator consists of two generation branches that model the person's appearance and shape information, respectively. Moreover, we propose two novel blocks to effectively transfer and update the person's shape and appearance embeddings in a crossing way to mutually improve each other, which has not been considered by any other existing GAN-based image generation work. Extensive experiments on two challenging datasets, i.e., Market-1501 and DeepFashion, demonstrate that the proposed XingGAN advances the state-of-the-art performance both in terms of objective quantitative scores and subjective visual realness. The source code and trained models are available at https://github.com/Ha0Tang/XingGAN.