OneHOI: Unifying Human-Object Interaction Generation and Editing
Paper • 2604.14062 • Published • 8
import torch
from diffusers import DiffusionPipeline
# switch to "mps" for apple devices
pipe = DiffusionPipeline.from_pretrained("jiuntian/OneHOI", dtype=torch.bfloat16, device_map="cuda")
prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k"
image = pipe(prompt).images[0]OneHOI: Unifying Human-Object Interaction Generation and Editing
OneHOI Project Page | Paper (CVPR 2026)
To use this pretrained models, see our Github repo at https://jiuntian.github.io/OneHOI/.
Citation
If you find this dataset useful for your research in HOI editing or image generation, please cite our CVPR 2026 paper:
@inproceedings{hoe2026onehoi,
title={OneHOI: Unifying Human-Object Interaction Generation and Editing},
author={Hoe, Jiun Tian and Hu, Weipeng and Jiang, Xudong and Tan, Yap-Peng and Chan, Chee Seng},
booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
year={2026}
}
@misc{hoe2025interactedit,
title={InteractEdit: Zero-Shot Editing of Human-Object Interactions in Images},
author={Jiun Tian Hoe and Weipeng Hu and Wei Zhou and Chao Xie and Ziwei Wang and Chee Seng Chan and Xudong Jiang and Yap-Peng Tan},
year={2025},
eprint={2503.09130},
archivePrefix={arXiv},
primaryClass={cs.GR},
url={https://arxiv.org/abs/2503.09130},
}
- Downloads last month
- 30