/pifuhd

High-Resolution 3D Human Digitization from A Single Image.

Primary LanguagePythonOtherNOASSERTION

report Open In Colab

This repository contains a pytorch implementation of "Multi-Level Pixel-Aligned Implicit Function for High-Resolution 3D Human Digitization".

Teaser Image

This codebase provides:

  • test code
  • visualization code

Demo on Google Colab

In case you don't have an environment to run PIFuHD, we offer Google Colab demo. You can also upload your own images and reconstruct 3D geometry together with visualization. Try our Colab demo using the following notebook:
Open In Colab

Requirements

  • Python 3
  • PyTorch tested on 1.1.0
  • json
  • PIL
  • skimage
  • tqdm
  • cv2

For visualization

  • trimesh with pyembree
  • PyOpenGL
  • freeglut (use sudo apt-get install freeglut3-dev for ubuntu users)
  • ffmpeg

Note: At least 8GB GPU memory is recommended to run PIFuHD model.

Download Pre-trained model

Run the following script to download the pretrained model. The checkpoint is saved under ./checkpoints/.

sh ./scripts/download_trained_model.sh

A Quick Testing

To process images under ./sample_images, run the following code:

sh ./scripts/demo.sh

The resulting obj files and rendering will be saved in ./results. You may use meshlab (http://www.meshlab.net/) to visualize the 3D mesh output (obj file).

Testing

  1. run the following script to get joints for each image for testing (joints are used for image cropping only.). Make sure you correctly set the location of OpenPose binary.
python apps/process_openpose.py -d {openpose_root_path} -i {path_of_images} -o {path_of_images}
  1. run the following script to run reconstruction code. Make sure to set --input_path to path_of_images, --out_path to where you want to dump out results, and --ckpt_path to the checkpoint. Note that unlike PIFu, PIFuHD doesn't require segmentation mask as input. But if you observe severe artifacts, you may try removing background with off-the-shelf tools such as removebg.
python -m apps.simple_test
  1. optionally, you can also remove artifacts by keeping only the biggest connected component from the mesh reconstruction with the following script. (Warning: the script will overwrite the original obj files.)
python apps/clean_mesh.py -f {path_of_objs}

Visualization

To render results with turn-table, run the following code. The rendered animation (.mp4) will be stored under {path_of_objs}.

python -m apps.render_turntable -f {path_of_objs} -ww {rendering_width} -hh {rendering_height} 
# add -g for geometry rendering. default is normal visualization.

Citation

@inproceedings{saito2020pifuhd,
  title={PIFuHD: Multi-Level Pixel-Aligned Implicit Function for High-Resolution 3D Human Digitization},
  author={Saito, Shunsuke and Simon, Tomas and Saragih, Jason and Joo, Hanbyul},
  booktitle={CVPR},
  year={2020}
}

Relevant Projects

Learning to Infer Implicit Surfaces without 3d Supervision (NeurIPS 2019)
Shichen Liu, Shunsuke Saito, Weikai Chen, Hao Li

We answer to the question of "how can we learn implicit function if we don't have 3D ground truth?"

PIFu: Pixel-Aligned Implicit Function for High-Resolution Clothed Human Digitization (ICCV 2019)
Shunsuke Saito*, Zeng Huang*, Ryota Natsume*, Shigeo Morishima, Angjoo Kanazawa, Hao Li

The original work of Pixel-Aligned Implicit Function for geometry and texture reconstruction, unifying sigle-view and multi-view methods.

SiCloPe: Silhouette-Based Clothed People (CVPR 2019, best paper finalist)
Ryota Natsume*, Shunsuke Saito*, Zeng Huang, Weikai Chen, Chongyang Ma, Hao Li, Shigeo Morishima

Our first attempt to reconstruct 3D clothed human body with texture from a single image!

Other Relevant Works

ARCH: Animatable Reconstruction of Clothed Humans (CVPR 2020)
Zeng Huang, Yuanlu Xu, Christoph Lassner, Hao Li, Tony Tung

Learning PIFu in canonical space for animatable avatar generation!

Robust 3D Self-portraits in Seconds (CVPR 2020)
Zhe Li, Tao Yu, Chuanyu Pan, Zerong Zheng, Yebin Liu

They extend PIFu to RGBD + introduce "PIFusion" utilizing PIFu reconstruction for non-rigid fusion.

Deep Volumetric Video from Very Sparse Multi-view Performance Capture (ECCV 2018)
Zeng Huang, Tianye Li, Weikai Chen, Yajie Zhao, Jun Xing, Chloe LeGendre, Linjie Luo, Chongyang Ma, Hao Li

Implict surface learning for sparse view human performance capture!

License

CC-BY-NC 4.0. See the LICENSE file.