"HDR-GS: 通过高斯点拼接以 1000 倍速度高效合成高动态范围的新视图" (NeurIPS 2024)
This is the official implementation of our NeurIPS 2024 paper "HDR-GS: Efficient High Dynamic Range Novel View Synthesis at 1000x Speed via Gaussian Splatting". We have run the SfM algorithm to recalibrate the data. If you find this repo useful, please give it a star ⭐ and consider citing our paper. Thank you.
Synthetic Datasets
Real Datasets
| Scene Bathroom | Scene Chair | Scene Diningroom |
|---|---|---|
| Scene Dog | Scene Sofa | Scene Sponza |
|---|---|---|
We recommend using Conda to set up the environment.
# cloning our repo
git clone https://github.com/caiyuanhao1998/HDR-GS --recursive
SET DISTUTILS_USE_SDK=1 # Windows only
# install the official environment of 3DGS
cd HDR-GS
conda env create --file environment.yml
conda activate hdr_gs
Download our recalibrated and reorganized datasets from Google Drive or Baidu Disk (code: cyh2). Then put the downloaded datasets into the folder data_hdr/ as
…Note: The original datasets are collected by HDR-NeRF. But the camera poses follow the normalized device coordinates, which are not suitable for 3DGS. Besides, HDR-NeRF does not provide the initial point clouds. So we use the Structure-from-Motion algorithm to recalibrate the camera poses and generate the initial point clouds. We also organize the datasets according to the description of HDR-NeRF, which is different from its official repo.
We write the code for directly loading the model to test and render spiral video. Please download our pre-trained weights bathroom from Google Drive or Baidu Disk (code: cyh2) and then put it into the folder pretrained_weights.
# For synthetic scenes
python3 train_synthetic.py --config config/bathroom.yaml --eval --gpu_id 0 --syn --load_path pretrained_weights/bathroom --test_onlyBesides, if you train a model with config: bathroom.yaml, you will get a profile as:
|--output
|--mlp
|--bathroom
|--exp-time
|--point_cloud
|interation_x
|--point_cloud.ply
|--tone_mapper.pth
...
|--test_set_vis
|--videos
|--cameras.json
|--cfg_args
|--input.ply
|--log.txtThen the --load_path should be "output/mlp/bathroom/exp-time/point_cloud/interation_x"
We provide training logs for your convenience to debug. Please download them from the Google Drive or Baidu Disk (code: cyh2).
You can run the .sh file by
# For synthetic scenes
bash train_synthetic.sh
# for real scenes
bash train_real.shOr you can directly train on specific scenes as
…
@inproceedings{hdr_gs,
title={HDR-GS: Efficient High Dynamic Range Novel View Synthesis at 1000x Speed via Gaussian Splatting},
author={Yuanhao Cai and Zihao Xiao and Yixun Liang and Minghan Qin and Yulun Zhang and Xiaokang Yang and Yaoyao Liu and Alan Yuille},
booktitle={NeurIPS},
year={2024}
}暂无开放 Issues,或尚未同步最近议题。