FashionPose: Text to Pose to Relight Image Generation for Personalized Fashion Visualization

TL;DR

FashionPose is a novel framework that generates personalized fashion images from text, predicting poses, synthesizing images, and relighting, enabling flexible and realistic virtual garment visualization for e-commerce.

Contribution

It introduces the first unified text-to-pose-to-relighting framework that replaces explicit pose annotations with text-driven conditioning for enhanced flexibility.

Findings

Achieves fine-grained pose synthesis

Enables efficient and consistent relighting

Provides a practical solution for virtual fashion display

Abstract

Realistic and controllable garment visualization is critical for fashion e-commerce, where users expect personalized previews under diverse poses and lighting conditions. Existing methods often rely on predefined poses, limiting semantic flexibility and illumination adaptability. To address this, we introduce FashionPose, the first unified text-to-pose-to-relighting generation framework. Given a natural language description, our method first predicts a 2D human pose, then employs a diffusion model to generate high-fidelity person images, and finally applies a lightweight relighting module, all guided by the same textual input. By replacing explicit pose annotations with text-driven conditioning, FashionPose enables accurate pose alignment, faithful garment rendering, and flexible lighting control. Experiments demonstrate fine-grained pose synthesis and efficient, consistent relighting, providing a practical solution for personalized virtual fashion display.