CraftsMan: High-fidelity Mesh Generation with 3D Native Generation and Interactive Geometry Refiner
Paper
β’
2405.14979
β’
Published
β’
19
Weiyu Li*1,2, Jiarui Liu*1,2, Hongyu Yan*1,2, Rui Chen1,2, Yixun Liang2,3, Xuelin Chen4, Ping Tan1,2, Xiaoxiao Long1,2
1HKUST, 2LightIllusions, 3HKUST(GZ), 4Tencent AI Lab
To use the model, please refer to the official repository for installation and usage instructions.
from craftsman import CraftsManPipeline
import torch
pipeline = CraftsManPipeline.from_pretrained("./ckpts/craftsman-v1-5", device="cuda:0", torch_dtype=torch.float32) # load from local ckpt
mesh = pipeline("https://pub-f9073a756ec645d692ce3d171c2e1232.r2.dev/data/werewolf.png").meshes[0]
mesh.export("werewolf.obj")
git clone https://github.com/wyysf-98/CraftsMan
cd CraftsMan
We provide an env_install.sh script file for setting up environment.
# step 1, create conda env
conda create -n CraftsMan python=3.10
conda activate CraftsMan
# step 2. install torch realated package
conda install -c pytorch pytorch=2.3.0 torchvision=0.18.0 cudatoolkit=11.8
# step 3. install other packages
pip install -r docker/requirements.txt
We have prepared a gradio demo for you to try out the model. You can run the following command to start the demo.
# std
python3 gradio.py
Then the demo can be accessed through the output link.
If you found this repository helpful, please cite our report:
@misc{li2024craftsman,
title = {CraftsMan: High-fidelity Mesh Generation with 3D Native Generation and Interactive Geometry Refiner},
author = {Weiyu Li and Jiarui Liu and Rui Chen and Yixun Liang and Xuelin Chen and Ping Tan and Xiaoxiao Long},
year = {2024},
archivePrefix = {arXiv preprint arXiv:2405.14979},
primaryClass = {cs.CG}
}