JiaxiongQ / NeuS-HSR

Looking Through the Glass: Neural Surface Reconstruction Against High Specular Reflections (CVPR 2023)

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

NeuS-HSR

This is the official repo for the implementation of "Looking Through the Glass: Neural Surface Reconstruction Against High Specular Reflections" paper (CVPR 2023). Arxiv Video

Synthetic Dataset

Google Drive only contains RGB images, camera parameters are from DTU Dataset.

Real-world Dataset

Google Drive

Comparison

Novel View Synthesis

Mirror case

Acknowlegement

Our code is built on NeuS project.

Our real-world dataset is constructed by following NeRF++ project.

Thanks for these great projects!

Citation

If you use our code or method in your work, please cite the following:

@InProceedings{Qiu_2023_CVPR,
author = {Qiu, Jiaxiong and Jiang, Peng-Tao and Zhu, Yifan and Yin, Ze-Xin and Cheng, Ming-Ming and Ren, Bo},
title = {Looking Through the Glass: Neural Surface Reconstruction Against High Specular Reflections},
booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
year = {2023}
}

About

Looking Through the Glass: Neural Surface Reconstruction Against High Specular Reflections (CVPR 2023)

License:MIT License


Languages

Language:Python 99.8%Language:Shell 0.2%