2d_gaussian_splatting.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "Before proceeding, please ensure that your notebook is connected to a GPU runtime."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Iu5nIIrEY1_n",
        "outputId": "8461bba3-1323-4dc9-bd1c-0bb46ede5122"
      },
      "outputs": [],
      "source": [
        "%cd /content\n",
        "!git clone https://github.com/atakan-topaloglu/2d-gaussian-splatting --recursive\n",
        "!pip install -q plyfile"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "P-XjteusY6YB",
        "outputId": "0d7ead53-7b6e-41ba-abce-58d68316c377"
      },
      "outputs": [],
      "source": [
        "%cd /content/2d-gaussian-splatting/"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VC9U4Lt1Y7o1",
        "outputId": "0891b956-4198-4ad0-c050-4887630a3b93"
      },
      "outputs": [],
      "source": [
        "import os\n",
        "os.environ['CUDA_HOME'] = '/usr/local/cuda' #ensure that the CUDA path is correct\n",
        "!pip install ninja\n",
        "!pip install submodules/simple-knn\n",
        "!pip install submodules/diff-surfel-rasterization"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Gp_mHmNfY8OH",
        "outputId": "7fe06939-e1b3-4abd-874d-06292532072b"
      },
      "outputs": [],
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mNbChlEEZ2ZP",
        "outputId": "2a8a9ac0-480b-4b0f-fc64-4d838af718e0"
      },
      "outputs": [],
      "source": [
        "%cd /content/2d-gaussian-splatting/"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aklgj8llF7wH"
      },
      "source": [
        "For Colab to access your COLMAP processed data, please copy the COLMAP output to your Google Drive and enter the full COLMAP source path (e.g. /content/drive/MyDrive/COLMAP_out)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UhHKVQ-Smo43",
        "outputId": "ef1d20aa-bb56-4248-bf1d-4d8f606d33b8"
      },
      "outputs": [],
      "source": [
        "COLMAP_PATH = str(input('Please enter COLMAP output path:'))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3GxTJBDKZ9xW",
        "outputId": "ea95cdb9-bbe6-4048-bebe-76537826a8e2"
      },
      "outputs": [],
      "source": [
        "!python train.py -s $COLMAP_PATH"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "Training is complete, to use render.py, please enter the desired model_path located in your Drive: /content/drive/MyDrive/2d_gaussian_splatting/output/{10 digit UUID}: (e.g: /content/drive/MyDrive/2d_gaussian_splatting/output/8d812792-1)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "MODEL_PATH = str(input('Please enter the model path:'))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Sn9uflTH5ut6",
        "outputId": "9e76f242-3a78-45b3-8d29-c8fcaf01c7db"
      },
      "outputs": [],
      "source": [
        "!pip3 install open3d==0.18.0 mediapy==1.1.2 lpips==0.1.4 scikit-image==0.21.0 tqdm==4.66.2 trimesh==4.3.2 opencv-python"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "T2inVQ3K49lS",
        "outputId": "77eaa496-61f1-42f8-8e75-1f3f445d698a"
      },
      "outputs": [],
      "source": [
        "!python render.py -m $MODEL_PATH -s $COLMAP_PATH"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "Training and rendering is complete, please navigate to your drive and see the model outputs under the directory \"2d_gaussian_splatting/\"."
      ]
    }
  ],
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}