TryOffAnyone: Tiled Cloth Generation from a Dressed Person

TL;DR

This paper introduces TryOffAnyone, a novel method using a fine-tuned StableDiffusion model to generate high-quality tiled garment images from photos of dressed individuals, enhancing virtual try-on and fashion recommendation systems.

Contribution

It presents a streamlined single-stage network with garment-specific masks and selective transformer training, achieving state-of-the-art results with reduced computational complexity.

Findings

Achieves high-fidelity tiled garment images from model photos.

Outperforms existing methods on benchmark datasets.

Reduces computational load while maintaining quality.

Abstract

The fashion industry is increasingly leveraging computer vision and deep learning technologies to enhance online shopping experiences and operational efficiencies. In this paper, we address the challenge of generating high-fidelity tiled garment images essential for personalized recommendations, outfit composition, and virtual try-on systems from photos of garments worn by models. Inspired by the success of Latent Diffusion Models (LDMs) in image-to-image translation, we propose a novel approach utilizing a fine-tuned StableDiffusion model. Our method features a streamlined single-stage network design, which integrates garmentspecific masks to isolate and process target clothing items effectively. By simplifying the network architecture through selective training of transformer blocks and removing unnecessary crossattention layers, we significantly reduce computational complexity while achieving state-of-the-art performance on benchmark datasets like VITON-HD. Experimental results demonstrate the effectiveness of our approach in producing high-quality tiled garment images for both full-body and half-body inputs. Code and model are available at: https://github.com/ixarchakos/try-off-anyone