hq0709 / depth-neus

Official code release for paper "Depth-NeuS: Neural Implicit Surfaces Learning for Multi-view Reconstruction Based on Depth Information Optimization"

Depth-NeuS

Contact author: click here

Data preparation

The data (from scannet) is organized as follows:

<scene_name>
|-- cameras_sphere.npz   # camera parameters
|-- image
    |-- 0000.png        # target image for each view
    |-- 0001.png
    ...
|-- depth
    |-- 0000.png        # target depth for each view
    |-- 0001.png
    ...
|-- pose
    |-- 0000.txt        # camera pose for each view
    |-- 0001.txt
    ...
|-- pred_normal
    |-- 0000.npz        # predicted normal for each view
    |-- 0001.npz
    ...
|-- xxx.ply		# GT mesh or point cloud from MVS
|-- trans_n2w.txt       # transformation matrix from normalized coordinates to world coordinates

Setup

pip install -r requirements.txt

Training

python ./exp_runner.py --mode train --conf ./confs/depth-neus.conf --gpu 0 --scene_name your_scene_name

Mesh Extraction

python exp_runner.py --mode validate_mesh --conf ./confs/depth-neus.conf --is_continue

Model Evaluation

python ./exp_evaluation.py --mode eval_3D_mesh_metrics

Citation

If you use the code, please cite the following paper:

@article{jiang2023depthneus,
      title={Depth-NeuS: Neural Implicit Surfaces Learning for Multi-view Reconstruction Based on Depth Information Optimization}, 
      author={Hanqi Jiang and Cheng Zeng and Runnan Chen and Shuai Liang and Yinhe Han and Yichao Gao and Conglin Wang},
      year={2023},
      eprint={2303.17088},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}

depth-neus's Issues

Comparison with the current depth-supervised surface reconstruction sota works

Hi authors, have you compared several existing sota works about depth-supervised surface reconstruction? Such as monosdf，neural rgb-d recon and so on.

How to train custom dataset?

Thanks for your great work!
I want to train my custom dataset, but how to make custom data and how to train custom datasets?
I cannot find trans_n2w.txt and pred_normal.txt in the scannet.
Thank you very much!