Structure-aware Person Image Generation with Pose Decomposition and Semantic Correlation
This work provides an incremental improvement for generating realistic person images, which is useful for applications in virtual try-on or animation.
This paper addresses pose-guided person image generation, aiming to transfer a person's appearance from a source pose to a target pose. The authors propose a structure-aware flow-based method that decomposes the human body into semantic parts and predicts flow fields separately for each part, achieving high-quality results under large pose discrepancy and outperforming state-of-the-art methods.
In this paper we tackle the problem of pose guided person image generation, which aims to transfer a person image from the source pose to a novel target pose while maintaining the source appearance. Given the inefficiency of standard CNNs in handling large spatial transformation, we propose a structure-aware flow based method for high-quality person image generation. Specifically, instead of learning the complex overall pose changes of human body, we decompose the human body into different semantic parts (e.g., head, torso, and legs) and apply different networks to predict the flow fields for these parts separately. Moreover, we carefully design the network modules to effectively capture the local and global semantic correlations of features within and among the human parts respectively. Extensive experimental results show that our method can generate high-quality results under large pose discrepancy and outperforms state-of-the-art methods in both qualitative and quantitative comparisons.