GraVITON: Graph based garment warping with attention guided inversion for Virtual-tryon

TL;DR

This paper presents GraVITON, a novel graph-based garment warping method that leverages attention-guided inversion and diffusion models to achieve highly realistic virtual try-on results with improved detail preservation.

Contribution

The paper introduces a graph-based warping technique combined with diffusion models and attention-guided inversion for enhanced virtual try-on realism.

Findings

State-of-the-art qualitative results on VITON-HD and Dresscode datasets.

Significant improvements in garment warping accuracy and texture preservation.

Effective occlusion-aware warping without holes or occlusions.

Abstract

Virtual try-on, a rapidly evolving field in computer vision, is transforming e-commerce by improving customer experiences through precise garment warping and seamless integration onto the human body. While existing methods such as TPS and flow address the garment warping but overlook the finer contextual details. In this paper, we introduce a novel graph based warping technique which emphasizes the value of context in garment flow. Our graph based warping module generates warped garment as well as a coarse person image, which is utilised by a simple refinement network to give a coarse virtual tryon image. The proposed work exploits latent diffusion model to generate the final tryon, treating garment transfer as an inpainting task. The diffusion model is conditioned with decoupled cross attention based inversion of visual and textual information. We introduce an occlusion aware warping constraint that generates dense warped garment, without any holes and occlusion. Our method, validated on VITON-HD and Dresscode datasets, showcases substantial state-of-the-art qualitative and quantitative results showing considerable improvement in garment warping, texture preservation, and overall realism.