3DGS 训练全流程记录
(NVIDIA P104-100)

RealityScan 2.2 对齐(手机照片 16 张)
  ├─ 导出 3.csv          → 相机位姿/内参/畸变(Internal/External camera parameters)
  ├─ 导出 images/        → COLMAP 配套图像(00000.png,已重命名)
  └─ 导出 sparse/0/      → COLMAP 稀疏重建(文本格式 txt)

本流程最终产物:3DGS 训练好的高斯点云(point_cloud.ply)+ 渲染图像

环境对齐(torch cu118 ↔ nvcc 11.8, numpy<2)
   → 获取源码(main + submodules,dr_aa 分支 + glm)
   → 编译三个 CUDA 扩展(MSVC + nvcc,绕过双向版本检查)
   → 数据准备(colmap txt→bin;图像与内参 0.5x 缩放)
   → 训练(7000 步验证)→ 渲染验证(PSNR)

# 显卡与架构
nvidia-smi --query-gpu=name,compute_cap,memory.total --format=csv
# 期望输出:NVIDIA P104-100, 6.1, 8192 MiB

# CUDA 工具链(nvcc 必须存在)
nvcc --version

# Python 依赖状态
python -c "import torch; print(torch.__version__, torch.cuda.is_available())"
python -c "import numpy; print(numpy.__version__)"

# MSVC 工具链(确认 cl.exe 存在即可,路径在 VS 安装目录)
ls "/c/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/"

# 系统 nvcc 是 11.8 → 安装 cu118 版 torch(必须用 PyTorch 官方源,默认源装的是 CPU 版!)
python -m pip install torch==2.4.1 torchvision==0.19.1 \
  --index-url https://download.pytorch.org/whl/cu118

# numpy 必须 <2(open-cv 等二进制扩展依赖 numpy 1.x)
python -m pip install "numpy==1.26.4"

# 验证
python -c "import torch; print(torch.__version__, torch.cuda.is_available())"   # 2.4.1+cu118 True
python -c "import numpy; print(numpy.__version__)"                              # 1.26.4

cd /d/temp

# 主仓库(main 分支)
curl -L --retry 10 -o gs.zip "https://ghfast.top/https://github.com/graphdeco-inria/gaussian-splatting/archive/refs/heads/main.zip"
unzip -q gs.zip && mv gaussian-splatting-main gaussian-splatting

# rasterizer:必须用 dr_aa 分支(主仓库代码调用 antialiasing 参数,main 分支无此接口)
curl -L --retry 10 -o dgr.zip "https://ghfast.top/https://github.com/graphdeco-inria/diff-gaussian-rasterization/archive/refs/heads/dr_aa.zip"
unzip -q dgr.zip
mkdir -p gaussian-splatting/submodules/diff-gaussian-rasterization
mv diff-gaussian-rasterization-dr_aa/* gaussian-splatting/submodules/diff-gaussian-rasterization/

# simple-knn 在 INRIA GitLab(不在 GitHub!)
curl -L --retry 10 -o sknn.zip "https://gitlab.inria.fr/bkerbl/simple-knn/-/archive/main/simple-knn-main.zip"
unzip -q sknn.zip
mkdir -p gaussian-splatting/submodules/simple-knn
mv simple-knn-main/* gaussian-splatting/submodules/simple-knn/

# fused-ssim(metrics 用)
curl -L --retry 10 -o fssim.zip "https://ghfast.top/https://github.com/rahul-goel/fused-ssim/archive/refs/heads/main.zip"
unzip -q fssim.zip
mkdir -p gaussian-splatting/submodules/fused-ssim
mv fused-ssim-main/* gaussian-splatting/submodules/fused-ssim/

# glm(rasterizer 的数学库依赖)
curl -L --retry 10 -o glm.zip "https://ghfast.top/https://github.com/g-truc/glm/archive/refs/tags/1.0.1.zip"
unzip -q glm.zip
mkdir -p gaussian-splatting/submodules/diff-gaussian-rasterization/third_party
cp -r glm-1.0.1/* gaussian-splatting/submodules/diff-gaussian-rasterization/third_party/glm/

"-allow-unsupported-compiler", "-D_ALLOW_COMPILER_AND_STL_VERSION_MISMATCH"

@echo off
call "C:\Program Files\Microsoft Visual Studio\2022\Community\VC\Auxiliary\Build\vcvars64.bat" >nul 2>&1
set PYTHONIOENCODING=utf-8
set DISTUTILS_USE_SDK=1
cd /d D:\temp\gaussian-splatting\submodules\diff-gaussian-rasterization
python setup.py install
cd /d D:\temp\gaussian-splatting\submodules\simple-knn
python setup.py install
cd /d D:\temp\gaussian-splatting\submodules\fused-ssim
python setup.py install

cd /d/temp/ReasonixChat/3/readme && cmd //c build_all.bat
cd /d/temp/gaussian-splatting
python -c "from diff_gaussian_rasterization import GaussianRasterizationSettings, GaussianRasterizer; print('dgr OK')"
python -c "import simple_knn; print('simple_knn OK')"
python -c "import fused_ssim; print('fused_ssim OK')"

# 1) txt → bin(16 相机 / 16 图像 / 23513 点)
cd /d/temp/ReasonixChat/3
python readme/colmap_txt2bin.py

# 2) 图像与内参同步缩放到 0.5x(适配 8GB 显存与训练速度)
#    产物:D:\temp\ReasonixChat\3\gs_data(images/ + sparse/0/*.bin)
python readme/prepare_gs_data.py 0.5

cd /d/temp/gaussian-splatting
PYTHONIOENCODING=utf-8 python train.py \
  -s "D:/temp/ReasonixChat/3/gs_data" \
  -m "D:/temp/ReasonixChat/3/gs_output" \
  --iterations 7000

Loading Training Cameras
Number of points at initialisation :  23513
Training progress: 100%|██████████| 7000/7000 [11:21<00:00, 9.49it/s, Loss=0.0155550]
[ITER 7000] Evaluating train: L1 0.009710725769400597 PSNR 36.59599494934082
[ITER 7000] Saving Gaussians
Training complete.

cd /d/temp/gaussian-splatting
PYTHONIOENCODING=utf-8 python render.py \
  -m "D:/temp/ReasonixChat/3/gs_output" \
  -s "D:/temp/ReasonixChat/3/gs_data" \
  --iteration 7000

import numpy as np
from PIL import Image
from pathlib import Path

r = Path("D:/temp/ReasonixChat/3/gs_output/train/ours_7000")
psnrs = []
for p in sorted((r / "renders").glob("*.png")):
    gt = np.asarray(Image.open(r / "gt" / p.name).convert("RGB"), dtype=np.float32) / 255.0
    pred = np.asarray(Image.open(p).convert("RGB"), dtype=np.float32) / 255.0
    mse = np.mean((pred - gt) ** 2)
    psnrs.append(10 * np.log10(1.0 / (mse + 1e-12)))
print(f"PSNR: mean={np.mean(psnrs):.2f} min={np.min(psnrs):.2f} max={np.max(psnrs):.2f}")

gs_output/
├── point_cloud/iteration_7000/point_cloud.ply   ← 训练好的高斯泼溅点云(核心产物)
├── input.ply                                     ← 稀疏点初始化
├── cameras.json / exposure.json                  ← 相机与曝光参数
├── train/ours_7000/{renders,gt}/                 ← 渲染图与真值
├── test/ours_7000/
└── events.out.tfevents.*                         ← TensorBoard 日志

cd /d/temp/gaussian-splatting
PYTHONIOENCODING=utf-8 python train.py \
  -s "D:/temp/ReasonixChat/3/gs_data" \
  -m "D:/temp/ReasonixChat/3/gs_output_full" \
  --iterations 30000

PYTHONIOENCODING=utf-8 ns-train splatfacto --data output --output-dir runs \
  --experiment-name myscan --max-num-iterations 30000 --vis tensorboard \
  nerfstudio-data --downscale-factor 2