---
license: mit
pipeline_tag: image-to-image
library_name: diffusers
---

# MCLD [![arXiv](https://img.shields.io/badge/arXiv-2503.15686-b31b1b.svg)](https://arxiv.org/abs/2503.15686)

> **Multi-focal Conditioned Latent Diffusion for Person Image Synthesis** <br>
> _Jiaqi Liu, Jichao Zhang, Paolo Rota, Nicu Sebe_<br>
> _Computer Vision and Pattern Recognition Conference (**CVPR**), 2025, Nashville, USA_

![qualitative](main_qualitative.png)

This repository contains the model described in [Multi-focal Conditioned Latent Diffusion for Person Image Synthesis](https://arxiv.org/abs/2503.15686).

Code: https://github.com/jqliu09/mcld

Editing Applications:
![editing](main_editing.png)