210 lines
6.7 KiB
Markdown
210 lines
6.7 KiB
Markdown
# Gaussian Haircut: Human Hair Reconstruction with Strand-Aligned 3D Gaussians
|
||
|
||
[**中文**](README_CN.md) | [**English**](README.md)
|
||
|
||
This repository contains the official implementation of Gaussian Haircut, a strand-based human hair reconstruction method from monocular video.
|
||
|
||
[**Paper**](https://arxiv.org/abs/2409.14778) | [**Project Page**](https://eth-ait.github.io/GaussianHaircut/)
|
||
|
||
## Overview
|
||
|
||
The reconstruction process includes the following main stages:
|
||
|
||
1. **Preprocessing Stage**
|
||
- Video frame extraction and organization
|
||
- COLMAP camera reconstruction
|
||
- Hair and body segmentation
|
||
- Image quality assessment and filtering
|
||
- Orientation map calculation
|
||
- Facial keypoint detection
|
||
- FLAME head model fitting
|
||
|
||
2. **Reconstruction Stage**
|
||
- 3D Gaussian reconstruction
|
||
- FLAME mesh fitting
|
||
- Scene cropping and optimization
|
||
- Hair strand reconstruction
|
||
|
||
3. **Visualization Stage**
|
||
- Export reconstructed strands
|
||
- Blender rendering visualization
|
||
- Generate result video
|
||
|
||
Expected output:
|
||
```
|
||
[your_scene_folder]/
|
||
├── raw.mp4 # Input video
|
||
├── 3d_gaussian_splatting/ # 3D Gaussian reconstruction results
|
||
├── flame_fitting/ # FLAME head model fitting results
|
||
├── strands_reconstruction/ # Hair strand reconstruction intermediate results
|
||
├── curves_reconstruction/ # Final hair strand results
|
||
└── visualization/ # Rendering results and video
|
||
```
|
||
|
||
## 入门指南
|
||
|
||
### Linux 平台
|
||
|
||
1. **安装 CUDA 11.8**
|
||
|
||
按照 https://developer.nvidia.com/cuda-11-8-0-download-archive 上的说明进行操作。
|
||
|
||
确保：
|
||
- PATH 包含 <CUDA_DIR>/bin
|
||
- LD_LIBRARY_PATH 包含 <CUDA_DIR>/lib64
|
||
|
||
该环境仅在此 CUDA 版本下进行了测试。
|
||
|
||
2. **安装 Blender 3.6** 以创建股线可视化
|
||
|
||
按照 https://www.blender.org/download/lts/3-6 上的说明进行操作。
|
||
|
||
3. **克隆仓库并运行安装脚本**
|
||
|
||
```bash
|
||
git clone git@github.com:eth-ait/GaussianHaircut.git
|
||
cd GaussianHaircut
|
||
chmod +x ./install.sh
|
||
./install.sh
|
||
```
|
||
|
||
### Windows 平台
|
||
|
||
1. **安装 CUDA 11.8**
|
||
- 从 https://developer.nvidia.com/cuda-11-8-0-download-archive 下载并安装
|
||
- 默认安装路径：C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8
|
||
- 确保CUDA版本与系统兼容
|
||
|
||
2. **安装 Blender 3.6**
|
||
- 从 https://www.blender.org/download/lts/3-6 下载并安装
|
||
- 默认安装路径：C:\Program Files\Blender Foundation\Blender 3.6
|
||
|
||
3. **安装 Visual Studio 2019 Build Tools**
|
||
- 从 https://visualstudio.microsoft.com/vs/older-downloads/ 下载并安装
|
||
- 选择"C++构建工具"工作负载
|
||
- 默认安装路径：C:\Program Files (x86)\Microsoft Visual Studio\2019\BuildTools
|
||
|
||
4. **安装 COLMAP**
|
||
- 从 https://github.com/colmap/colmap/releases 下载并安装
|
||
- 下载CUDA版本的COLMAP (例如：COLMAP-3.8-windows-cuda.zip)
|
||
- 解压到不含空格的路径 (例如：C:\COLMAP)
|
||
- 将COLMAP目录添加到系统PATH:
|
||
1. 打开"系统属性" > "环境变量"
|
||
2. 在"系统变量"中找到"Path"
|
||
3. 点击"编辑" > "新建"
|
||
4. 添加COLMAP目录路径
|
||
5. 点击"确定"保存
|
||
- 重启终端使PATH生效
|
||
|
||
5. **安装 7-Zip**
|
||
- 从 https://7-zip.org/ 下载并安装
|
||
- 将7-Zip安装目录添加到系统PATH:
|
||
1. 打开"系统属性" > "环境变量"
|
||
2. 在"系统变量"中找到"Path"
|
||
3. 点击"编辑" > "新建"
|
||
4. 添加7-Zip安装目录(默认为C:\Program Files\7-Zip)
|
||
5. 点击"确定"保存
|
||
- 重启终端使PATH生效
|
||
|
||
6. **下载预训练模型和资源**
|
||
```cmd
|
||
git clone https://github.com/Jeffreytsai1004/GaussianHairCut
|
||
cd GaussianHairCut
|
||
# 在PowerShell中运行:
|
||
# 脚本会自动安装gdown并下载所需资源
|
||
.\download_resource.bat
|
||
```
|
||
注意：
|
||
- 下载过程可能需要几分钟到几十分钟，取决于网络速度
|
||
- 如果下载失败，可以重新运行脚本
|
||
- 确保有稳定的网络连接
|
||
|
||
6. **克隆仓库并运行安装脚本**
|
||
```cmd
|
||
git clone https://github.com/Jeffreytsai1004/GaussianHairCut
|
||
cd GaussianHairCut
|
||
# 先下载所需要的资源
|
||
.\download_resource.bat
|
||
# 运行安装脚本
|
||
.\install.bat
|
||
# 运行重建脚本
|
||
.\run.bat
|
||
```
|
||
|
||
## 重建
|
||
|
||
1. **录制单目视频**
|
||
|
||
- 参考项目页面上的示例视频
|
||
- 录制要求：
|
||
* 拍摄对象应缓慢转动头部，确保捕捉到所有角度
|
||
* 保持头发和面部清晰可见
|
||
* 避免快速移动导致的运动模糊
|
||
* 保持光照条件稳定
|
||
* 建议视频长度：10-20秒
|
||
* 建议分辨率：1920x1080或更高
|
||
|
||
2. **设置重建场景的目录**
|
||
|
||
创建场景目录结构：
|
||
```
|
||
[your_scene_folder]/
|
||
└── raw.mp4 # 您录制的视频文件
|
||
```
|
||
|
||
注意：
|
||
- DATA_PATH 应指向包含 raw.mp4 的目录
|
||
- 目录路径不应包含空格或特殊字符
|
||
- 脚本会在此目录下自动创建所需的子目录
|
||
- 确保有足够的磁盘空间(建议至少20GB)
|
||
|
||
3. **运行脚本**
|
||
|
||
Linux:
|
||
```bash
|
||
export PROJECT_DIR="[/path/to/]GaussianHaircut"
|
||
export BLENDER_DIR="[/path/to/blender/folder/]blender"
|
||
DATA_PATH="[path/to/scene/folder]" ./run.sh
|
||
```
|
||
|
||
Windows:
|
||
```cmd
|
||
# 在CMD中运行:
|
||
set PROJECT_DIR=[path\to\]GaussianHaircut
|
||
set DATA_PATH=[path\to\scene\folder]
|
||
run.bat
|
||
|
||
# 或在PowerShell中运行:
|
||
set PROJECT_DIR=[path\to\]GaussianHaircut
|
||
set DATA_PATH=[path\to\scene\folder]
|
||
.\run.bat
|
||
```
|
||
|
||
该脚本执行数据预处理、重建以及使用 Blender 进行最终可视化。使用 Tensorboard 查看中间可视化结果。
|
||
|
||
## 许可证
|
||
|
||
此代码基于 3D Gaussian Splatting 项目。有关条款和条件，请参阅 LICENSE_3DGS。其余代码根据 CC BY-NC-SA 4.0 分发。
|
||
|
||
如果此代码对您的项目有帮助，请引用以下论文。
|
||
|
||
## 引用
|
||
|
||
```
|
||
@inproceedings{zakharov2024gh,
|
||
title = {Human Hair Reconstruction with Strand-Aligned 3D Gaussians},
|
||
author = {Zakharov, Egor and Sklyarova, Vanessa and Black, Michael J and Nam, Giljoo and Thies, Justus and Hilliges, Otmar},
|
||
booktitle = {European Conference of Computer Vision (ECCV)},
|
||
year = {2024}
|
||
}
|
||
```
|
||
|
||
## 链接
|
||
|
||
- [3D Gaussian Splatting](https://github.com/graphdeco-inria/gaussian-splatting)
|
||
- [Neural Haircut](https://github.com/SamsungLabs/NeuralHaircut): FLAME 拟合管线、股线先验和发型扩散先验
|
||
- [HAAR](https://github.com/Vanessik/HAAR): 头发上采样
|
||
- [Matte-Anything](https://github.com/hustvl/Matte-Anything): 头发和身体分割
|
||
- [PIXIE](https://github.com/yfeng95/PIXIE): FLAME 拟合的初始化
|
||
- [Face-Alignment](https://github.com/1adrianb/face-alignment), [OpenPose](https://github.com/CMU-Perceptual-Computing-Lab/openpose): 用于 FLAME 拟合的关键点检测
|