PersonaHOI: Effortlessly Improving Personalized Face with Human-Object Interaction Generation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Hu, Xinting, Wang, Haoran, Lenssen, Jan Eric, Schiele, Bernt
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929669898502144
author Hu, Xinting
Wang, Haoran
Lenssen, Jan Eric
Schiele, Bernt
author_facet Hu, Xinting
Wang, Haoran
Lenssen, Jan Eric
Schiele, Bernt
contents We introduce PersonaHOI, a training- and tuning-free framework that fuses a general StableDiffusion model with a personalized face diffusion (PFD) model to generate identity-consistent human-object interaction (HOI) images. While existing PFD models have advanced significantly, they often overemphasize facial features at the expense of full-body coherence, PersonaHOI introduces an additional StableDiffusion (SD) branch guided by HOI-oriented text inputs. By incorporating cross-attention constraints in the PFD branch and spatial merging at both latent and residual levels, PersonaHOI preserves personalized facial details while ensuring interactive non-facial regions. Experiments, validated by a novel interaction alignment metric, demonstrate the superior realism and scalability of PersonaHOI, establishing a new standard for practical personalized face with HOI generation. Our code will be available at https://github.com/JoyHuYY1412/PersonaHOI
format Preprint
id arxiv_https___arxiv_org_abs_2501_05823
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle PersonaHOI: Effortlessly Improving Personalized Face with Human-Object Interaction Generation
Hu, Xinting
Wang, Haoran
Lenssen, Jan Eric
Schiele, Bernt
Computer Vision and Pattern Recognition
We introduce PersonaHOI, a training- and tuning-free framework that fuses a general StableDiffusion model with a personalized face diffusion (PFD) model to generate identity-consistent human-object interaction (HOI) images. While existing PFD models have advanced significantly, they often overemphasize facial features at the expense of full-body coherence, PersonaHOI introduces an additional StableDiffusion (SD) branch guided by HOI-oriented text inputs. By incorporating cross-attention constraints in the PFD branch and spatial merging at both latent and residual levels, PersonaHOI preserves personalized facial details while ensuring interactive non-facial regions. Experiments, validated by a novel interaction alignment metric, demonstrate the superior realism and scalability of PersonaHOI, establishing a new standard for practical personalized face with HOI generation. Our code will be available at https://github.com/JoyHuYY1412/PersonaHOI
title PersonaHOI: Effortlessly Improving Personalized Face with Human-Object Interaction Generation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2501.05823