FW-VTON: Flattening-and-Warping for Person-to-Person Virtual Try-on

2025-07-21

AI Summary

This paper addresses the novel task of person-to-person virtual try-on—generating photorealistic synthesis of a target person wearing a garment from an image of another person wearing it, given only the target person’s image and the source garment image. We propose the first end-to-end, three-stage Flattening-and-Warping framework: (1) segmenting and flattening the garment from the source image into a canonical 2D layout; (2) pose-guided warping to align the flattened garment with the target person’s body geometry; and (3) detail-preserving fusion and adversarial refinement for enhanced realism. To support this task, we introduce the first high-quality person-to-person virtual try-on benchmark dataset. Our method significantly outperforms existing approaches in garment segmentation accuracy, warping robustness under diverse poses, and overall visual fidelity, achieving state-of-the-art performance across multiple quantitative metrics and qualitative evaluations.

Technology Category

• Computer Vision: Computational Photography, Image & Video Synthesis
• Humans and AI: Game Design — Virtual Humans, NPCs and Autonomous Characters
• Cognitive Modeling & Cognitive Systems: Simulating Human Behavior

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Traditional virtual try-on methods primarily focus on the garment-to-person try-on task, which requires flat garment representations. In contrast, this paper introduces a novel approach to the person-to-person try-on task. Unlike the garment-to-person try-on task, the person-to-person task only involves two input images: one depicting the target person and the other showing the garment worn by a different individual. The goal is to generate a realistic combination of the target person with the desired garment. To this end, we propose Flattening-and-Warping Virtual Try-On ( extbf{FW-VTON}), a method that operates in three stages: (1) extracting the flattened garment image from the source image; (2) warping the garment to align with the target pose; and (3) integrating the warped garment seamlessly onto the target person. To overcome the challenges posed by the lack of high-quality datasets for this task, we introduce a new dataset specifically designed for person-to-person try-on scenarios. Experimental evaluations demonstrate that FW-VTON achieves state-of-the-art performance, with superior results in both qualitative and quantitative assessments, and also excels in garment extraction subtasks.

Problem

• Develops person-to-person virtual try-on using two images
• Proposes FW-VTON for garment flattening, warping, and integration
• Introduces new dataset for person-to-person try-on scenarios

Innovation

• Extracts flattened garment from source image
• Warps garment to match target pose
• Integrates garment seamlessly onto target