Pixel2Mesh

This repository contains the TensorFlow implementation for the following paper

Pixel2Mesh: Generating 3D Mesh Models from Single RGB Images (ECCV2018)

Nanyang Wang*, Yinda Zhang*, Zhuwen Li*, Yanwei Fu, Wei Liu, Yu-Gang Jiang.

The code is based on the gcn framework. For Chamfer losses, we have included the cuda implementations of Fan et. al.

Citation

If you use this code for your research, please consider citing:

@inProceedings{wang2018pixel2mesh,
  title={Pixel2Mesh: Generating 3D Mesh Models from Single RGB Images},
  author={Nanyang Wang and Yinda Zhang and Zhuwen Li and Yanwei Fu and Wei Liu and Yu-Gang Jiang},
  booktitle={ECCV},
  year={2018}
}

Project Page

The project page is available at http://bigvid.fudan.edu.cn/pixel2mesh

Dependencies

Requirements:

Python2.7+ with Numpy and opencv-python
Tensorflow (version 1.0+)
TFLearn

Our code has been tested with Python 2.7, TensorFlow 1.3.0, TFLearn 0.3.2, CUDA 8.0 on Ubuntu 14.04.

Installation

git clone https://github.com/nywang16/Pixel2Mesh.git
cd Pixel2Mesh
python setup.py install

Running the demo

First, you need download pre-trained models.

version 1: http://bigvid.fudan.edu.cn/pixel2mesh/eccv2018/checkpoint.zip
version 2: We will release an improved model in a few days !!

cd pixel2mesh/utils
wget http://bigvid.fudan.edu.cn/pixel2mesh/eccv2018/checkpoint.zip
unzip checkpoint.zip

Reconstructing shapes

python demo.py --image utils/examples/plane.png

Run the demo code and the output mesh file is saved in utils/examples/plane.obj

Input image, output mesh

Dataset

We used the ShapeNet dataset for 3D models, and rendered views from 3D-R2N2:
When using the provided data make sure to respect the shapenet license.

Below is the complete set of training data. Download it into the utils/ folder. https://drive.google.com/file/d/1Z8gt4HdPujBNFABYrthhau9VZW10WWYe/view?usp=sharing

cd pixel2mesh/utils
tar -xzf ShapeNetTrain.tar

The training/testing split can be found in utils/train_list.txt and utils/test_list.txt

The file is named in syntheticID_modelID_renderID.dat format, and the data processing script will be released soon.

Each .dat file in the provided data contain:

The rendered image from 3D-R2N2. We resized it to 224x224 and made the background white.
The sampled point cloud (with vertex normal) from ShapeNet. We transformed it to corresponding coordinates in camera coordinate based on camera parameters from the Rendering Dataset.

Input image, ground truth point cloud.

Training

python train.py

You can change the training data, learning rate and other parameters by editing train.py

Statement

This software is for research purpose only.
Please contact us for the licence of commercial purposes. All rights are preserved.

Contact

Nanyang Wang (nywang16 AT 163.com)

Yinda Zhang (yindaz AT cs.princeton.edu)

Zhuwen Li (lzhuwen AT gmail.com)

Yanwei Fu (yanweifu AT fudan.edu.cn)

Yu-Gang Jiang (ygj AT fudan.edu.cn)

Any comments or suggestions are welcome !

liufeng1990/Pixel2Mesh