Pipeline_3D_Photo_Inpainting.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Pipeline-3D-Photo-Inpainting.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IY-s3ZpHBeAU"
      },
      "source": [
        "**[CVPR 2020] 3D Photography using Context-aware Layered Depth Inpainting**\n",
        "\n",
        "[project website](https://shihmengli.github.io/3D-Photo-Inpainting/)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LPQ0EWxgvMIc"
      },
      "source": [
        "- **Prepare environment**\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5o-EIMeaghU0",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "2c07887c-a92b-4021-900d-9294f2e03ae8"
      },
      "source": [
        "!pip3 install opencv-python==4.2.0.32\n",
        "!pip3 install vispy==0.6.4\n",
        "!pip3 install moviepy==1.0.2\n",
        "!pip3 install transforms3d==0.3.1\n",
        "!pip3 install networkx==2.3\n",
        "!sudo apt install sed"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Collecting opencv-python==4.2.0.32\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/34/a3/403dbaef909fee9f9f6a8eaff51d44085a14e5bb1a1ff7257117d744986a/opencv_python-4.2.0.32-cp37-cp37m-manylinux1_x86_64.whl (28.2MB)\n",
            "\u001b[K     |████████████████████████████████| 28.2MB 107kB/s \n",
            "\u001b[?25hRequirement already satisfied: numpy>=1.14.5 in /usr/local/lib/python3.7/dist-packages (from opencv-python==4.2.0.32) (1.19.5)\n",
            "\u001b[31mERROR: albumentations 0.1.12 has requirement imgaug<0.2.7,>=0.2.5, but you'll have imgaug 0.2.9 which is incompatible.\u001b[0m\n",
            "Installing collected packages: opencv-python\n",
            "  Found existing installation: opencv-python 4.1.2.30\n",
            "    Uninstalling opencv-python-4.1.2.30:\n",
            "      Successfully uninstalled opencv-python-4.1.2.30\n",
            "Successfully installed opencv-python-4.2.0.32\n",
            "Collecting vispy==0.6.4\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/5f/7e/9e24a986e8fe2ee94588f806e1c4579c8c285e7e42b70189d79db4132e07/vispy-0.6.4-cp37-cp37m-manylinux2010_x86_64.whl (2.3MB)\n",
            "\u001b[K     |████████████████████████████████| 2.3MB 4.0MB/s \n",
            "\u001b[?25hRequirement already satisfied: numpy in /usr/local/lib/python3.7/dist-packages (from vispy==0.6.4) (1.19.5)\n",
            "Collecting freetype-py\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/40/b5/56751e13f8b4a42f52c1b85ecce1446f83245190d820b42748eb8574ae43/freetype_py-2.2.0-py3-none-manylinux1_x86_64.whl (890kB)\n",
            "\u001b[K     |████████████████████████████████| 890kB 18.8MB/s \n",
            "\u001b[?25hInstalling collected packages: freetype-py, vispy\n",
            "Successfully installed freetype-py-2.2.0 vispy-0.6.4\n",
            "Collecting moviepy==1.0.2\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/71/64/e7f0876716d5c35612bf8e9e49299d0f491285f10f36cbb481556ab0062b/moviepy-1.0.2.tar.gz (7.8MB)\n",
            "\u001b[K     |████████████████████████████████| 7.9MB 3.7MB/s \n",
            "\u001b[?25hRequirement already satisfied: decorator<5.0,>=4.0.2 in /usr/local/lib/python3.7/dist-packages (from moviepy==1.0.2) (4.4.2)\n",
            "Requirement already satisfied: tqdm<5.0,>=4.11.2 in /usr/local/lib/python3.7/dist-packages (from moviepy==1.0.2) (4.41.1)\n",
            "Requirement already satisfied: requests<3.0,>=2.8.1 in /usr/local/lib/python3.7/dist-packages (from moviepy==1.0.2) (2.23.0)\n",
            "Collecting proglog<=1.0.0\n",
            "  Downloading https://files.pythonhosted.org/packages/fe/ab/4cb19b578e1364c0b2d6efd6521a8b4b4e5c4ae6528041d31a2a951dd991/proglog-0.1.9.tar.gz\n",
            "Requirement already satisfied: numpy>=1.17.3 in /usr/local/lib/python3.7/dist-packages (from moviepy==1.0.2) (1.19.5)\n",
            "Collecting imageio<3.0,>=2.5\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/6e/57/5d899fae74c1752f52869b613a8210a2480e1a69688e65df6cb26117d45d/imageio-2.9.0-py3-none-any.whl (3.3MB)\n",
            "\u001b[K     |████████████████████████████████| 3.3MB 37.1MB/s \n",
            "\u001b[?25hCollecting imageio_ffmpeg>=0.2.0\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/89/0f/4b49476d185a273163fa648eaf1e7d4190661d1bbf37ec2975b84df9de02/imageio_ffmpeg-0.4.3-py3-none-manylinux2010_x86_64.whl (26.9MB)\n",
            "\u001b[K     |████████████████████████████████| 26.9MB 118kB/s \n",
            "\u001b[?25hRequirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.7/dist-packages (from requests<3.0,>=2.8.1->moviepy==1.0.2) (2020.12.5)\n",
            "Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.7/dist-packages (from requests<3.0,>=2.8.1->moviepy==1.0.2) (2.10)\n",
            "Requirement already satisfied: urllib3!=1.25.0,!=1.25.1,<1.26,>=1.21.1 in /usr/local/lib/python3.7/dist-packages (from requests<3.0,>=2.8.1->moviepy==1.0.2) (1.24.3)\n",
            "Requirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.7/dist-packages (from requests<3.0,>=2.8.1->moviepy==1.0.2) (3.0.4)\n",
            "Requirement already satisfied: pillow in /usr/local/lib/python3.7/dist-packages (from imageio<3.0,>=2.5->moviepy==1.0.2) (7.0.0)\n",
            "Building wheels for collected packages: moviepy, proglog\n",
            "  Building wheel for moviepy (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for moviepy: filename=moviepy-1.0.2-cp37-none-any.whl size=110730 sha256=83d77fce3e46c94da42867cc48ee290e4b1f113303e2afb98661bf44bd1e585c\n",
            "  Stored in directory: /root/.cache/pip/wheels/20/49/a9/770a9a22e7a020d796ab3784d2dc4839e38450a1ac87e6106f\n",
            "  Building wheel for proglog (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for proglog: filename=proglog-0.1.9-cp37-none-any.whl size=6148 sha256=775c1b25c1d18a01ad0cf8a8d6240bbcff173dc089cbba65066c8e2d96c93f26\n",
            "  Stored in directory: /root/.cache/pip/wheels/65/56/60/1d0306a8d90b188af393c1812ddb502a8821b70917f82dcc00\n",
            "Successfully built moviepy proglog\n",
            "\u001b[31mERROR: albumentations 0.1.12 has requirement imgaug<0.2.7,>=0.2.5, but you'll have imgaug 0.2.9 which is incompatible.\u001b[0m\n",
            "Installing collected packages: proglog, imageio, imageio-ffmpeg, moviepy\n",
            "  Found existing installation: imageio 2.4.1\n",
            "    Uninstalling imageio-2.4.1:\n",
            "      Successfully uninstalled imageio-2.4.1\n",
            "  Found existing installation: moviepy 0.2.3.5\n",
            "    Uninstalling moviepy-0.2.3.5:\n",
            "      Successfully uninstalled moviepy-0.2.3.5\n",
            "Successfully installed imageio-2.9.0 imageio-ffmpeg-0.4.3 moviepy-1.0.2 proglog-0.1.9\n",
            "Collecting transforms3d==0.3.1\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/b5/f7/e85809168a548a854d7c1331560c27b4f5381698d29c12e57759192b2bc1/transforms3d-0.3.1.tar.gz (62kB)\n",
            "\u001b[K     |████████████████████████████████| 71kB 2.5MB/s \n",
            "\u001b[?25hBuilding wheels for collected packages: transforms3d\n",
            "  Building wheel for transforms3d (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for transforms3d: filename=transforms3d-0.3.1-cp37-none-any.whl size=59374 sha256=c501e19b9c4f6ac836c995752bbada6ca8c0ca01f4b8f3c4d9caacd78a9afc73\n",
            "  Stored in directory: /root/.cache/pip/wheels/3d/3c/84/28d36677f3c760c048bd02b5a547ea0c4027770cc9cdb9af1e\n",
            "Successfully built transforms3d\n",
            "Installing collected packages: transforms3d\n",
            "Successfully installed transforms3d-0.3.1\n",
            "Collecting networkx==2.3\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/85/08/f20aef11d4c343b557e5de6b9548761811eb16e438cee3d32b1c66c8566b/networkx-2.3.zip (1.7MB)\n",
            "\u001b[K     |████████████████████████████████| 1.8MB 4.1MB/s \n",
            "\u001b[?25hRequirement already satisfied: decorator>=4.3.0 in /usr/local/lib/python3.7/dist-packages (from networkx==2.3) (4.4.2)\n",
            "Building wheels for collected packages: networkx\n",
            "  Building wheel for networkx (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for networkx: filename=networkx-2.3-py2.py3-none-any.whl size=1556408 sha256=66af787da41ed4b477291eaf48b92e5e957b22463f2ef8e47f0bff8f1e648f9d\n",
            "  Stored in directory: /root/.cache/pip/wheels/de/63/64/3699be2a9d0ccdb37c7f16329acf3863fd76eda58c39c737af\n",
            "Successfully built networkx\n",
            "\u001b[31mERROR: albumentations 0.1.12 has requirement imgaug<0.2.7,>=0.2.5, but you'll have imgaug 0.2.9 which is incompatible.\u001b[0m\n",
            "Installing collected packages: networkx\n",
            "  Found existing installation: networkx 2.5\n",
            "    Uninstalling networkx-2.5:\n",
            "      Successfully uninstalled networkx-2.5\n",
            "Successfully installed networkx-2.3\n",
            "Reading package lists... Done\n",
            "Building dependency tree       \n",
            "Reading state information... Done\n",
            "sed is already the newest version (4.4-2).\n",
            "0 upgraded, 0 newly installed, 0 to remove and 30 not upgraded.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "c-g7AeLMvY0R"
      },
      "source": [
        "- **Download script and pretrained model**\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cOFIBkWrBlNM",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "fe5c812f-5287-49a6-865d-ceb82ba3b22d"
      },
      "source": [
        "%cd /content/\n",
        "!git clone https://github.com/vt-vl-lab/3d-photo-inpainting.git\n",
        "%cd 3d-photo-inpainting\n",
        "!sh download.sh"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/content\n",
            "Cloning into '3d-photo-inpainting'...\n",
            "remote: Enumerating objects: 350, done.\u001b[K\n",
            "remote: Total 350 (delta 0), reused 0 (delta 0), pack-reused 350\u001b[K\n",
            "Receiving objects: 100% (350/350), 130.13 MiB | 26.19 MiB/s, done.\n",
            "Resolving deltas: 100% (192/192), done.\n",
            "/content/3d-photo-inpainting\n",
            "downloading from filebox ...\n",
            "--2021-03-25 16:08:37--  https://filebox.ece.vt.edu/~jbhuang/project/3DPhoto/model/color-model.pth\n",
            "Resolving filebox.ece.vt.edu (filebox.ece.vt.edu)... 128.173.88.43\n",
            "Connecting to filebox.ece.vt.edu (filebox.ece.vt.edu)|128.173.88.43|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 206331633 (197M)\n",
            "Saving to: ‘color-model.pth’\n",
            "\n",
            "color-model.pth     100%[===================>] 196.77M  14.9MB/s    in 17s     \n",
            "\n",
            "2021-03-25 16:08:55 (11.6 MB/s) - ‘color-model.pth’ saved [206331633/206331633]\n",
            "\n",
            "--2021-03-25 16:08:55--  https://filebox.ece.vt.edu/~jbhuang/project/3DPhoto/model/depth-model.pth\n",
            "Resolving filebox.ece.vt.edu (filebox.ece.vt.edu)... 128.173.88.43\n",
            "Connecting to filebox.ece.vt.edu (filebox.ece.vt.edu)|128.173.88.43|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 206272258 (197M)\n",
            "Saving to: ‘depth-model.pth’\n",
            "\n",
            "depth-model.pth     100%[===================>] 196.72M  12.9MB/s    in 26s     \n",
            "\n",
            "2021-03-25 16:09:22 (7.45 MB/s) - ‘depth-model.pth’ saved [206272258/206272258]\n",
            "\n",
            "--2021-03-25 16:09:22--  https://filebox.ece.vt.edu/~jbhuang/project/3DPhoto/model/edge-model.pth\n",
            "Resolving filebox.ece.vt.edu (filebox.ece.vt.edu)... 128.173.88.43\n",
            "Connecting to filebox.ece.vt.edu (filebox.ece.vt.edu)|128.173.88.43|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 45974122 (44M)\n",
            "Saving to: ‘edge-model.pth’\n",
            "\n",
            "edge-model.pth      100%[===================>]  43.84M  5.80MB/s    in 14s     \n",
            "\n",
            "2021-03-25 16:09:37 (3.09 MB/s) - ‘edge-model.pth’ saved [45974122/45974122]\n",
            "\n",
            "--2021-03-25 16:09:37--  https://filebox.ece.vt.edu/~jbhuang/project/3DPhoto/model/model.pt\n",
            "Resolving filebox.ece.vt.edu (filebox.ece.vt.edu)... 128.173.88.43\n",
            "Connecting to filebox.ece.vt.edu (filebox.ece.vt.edu)|128.173.88.43|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 149751722 (143M)\n",
            "Saving to: ‘model.pt’\n",
            "\n",
            "model.pt            100%[===================>] 142.81M  15.3MB/s    in 14s     \n",
            "\n",
            "2021-03-25 16:09:52 (10.5 MB/s) - ‘model.pt’ saved [149751722/149751722]\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "D_Fg_aMeDnmd"
      },
      "source": [
        "- **Switch off off-screen rendering**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "S9TmdOBRDXcV"
      },
      "source": [
        "!sed -i 's/offscreen_rendering: True/offscreen_rendering: False/g' argument.yml"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XJzKbMi3ODtF"
      },
      "source": [
        "Change argument.yml: require_midas: True to require_midas: False"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cCN8L7GiDpZA"
      },
      "source": [
        "- **Please upload `.jpg` files to `/content/3d-photo-inpainting-master/image/`**\n",
        "  - You can run this step multiple times to upload multiple `.jpg` files."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "n-Kl6fo1oows",
        "colab": {
          "resources": {
            "http://localhost:8080/nbextensions/google.colab/files.js": {
              "data": "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",
              "ok": true,
              "headers": [
                [
                  "content-type",
                  "application/javascript"
                ]
              ],
              "status": 200,
              "status_text": "OK"
            }
          },
          "base_uri": "https://localhost:8080/",
          "height": 74
        },
        "outputId": "b49642e4-22d3-42b6-9bf3-274d675a121c"
      },
      "source": [
        "%cd image\n",
        "from google.colab import files\n",
        "uploaded = files.upload()\n",
        "for fn in uploaded.keys():\n",
        "  print('User uploaded file \"{name}\" with length {length} bytes'.format(\n",
        "      name=fn, length=len(uploaded[fn])))\n",
        "%cd .."
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/content/3d-photo-inpainting/image\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "\n",
              "     <input type=\"file\" id=\"files-d1b6edc8-1d7d-415e-8a3d-9e632eb4a9ee\" name=\"files[]\" multiple disabled\n",
              "        style=\"border:none\" />\n",
              "     <output id=\"result-d1b6edc8-1d7d-415e-8a3d-9e632eb4a9ee\">\n",
              "      Upload widget is only available when the cell has been executed in the\n",
              "      current browser session. Please rerun this cell to enable.\n",
              "      </output>\n",
              "      <script src=\"/nbextensions/google.colab/files.js\"></script> "
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "/content/3d-photo-inpainting\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0sExX8x6IKtP",
        "outputId": "f6398c15-8fb4-4a83-848a-bebfdb6d8416"
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Mounted at /content/drive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-u2nGVVaH1vm"
      },
      "source": [
        "import torch.nn as nn\n",
        "class MobileNetSkipConcat(nn.Module):\n",
        "    def __init__(self, output_size, pretrained=True):\n",
        "\n",
        "        super(MobileNetSkipConcat, self).__init__()\n",
        "        self.output_size = output_size\n",
        "        mobilenet = MobileNet1()\n",
        "        if pretrained:\n",
        "            pretrained_path = os.path.join('imagenet', 'results', 'imagenet.arch=mobilenet.lr=0.1.bs=256', 'model_best.pth.tar')\n",
        "            checkpoint = torch.load(pretrained_path)\n",
        "            state_dict = checkpoint['state_dict']\n",
        "\n",
        "            from collections import OrderedDict\n",
        "            new_state_dict = OrderedDict()\n",
        "            for k, v in state_dict.items():\n",
        "                name = k[7:] # remove `module.`\n",
        "                new_state_dict[name] = v\n",
        "            mobilenet.load_state_dict(new_state_dict)\n",
        "        else:\n",
        "            mobilenet.apply(weights_init)\n",
        "\n",
        "        for i in range(14):\n",
        "            setattr( self, 'conv{}'.format(i), mobilenet.model[i])\n",
        "\n",
        "        kernel_size = 5\n",
        "        # self.decode_conv1 = conv(1024, 512, kernel_size)\n",
        "        # self.decode_conv2 = conv(512, 256, kernel_size)\n",
        "        # self.decode_conv3 = conv(256, 128, kernel_size)\n",
        "        # self.decode_conv4 = conv(128, 64, kernel_size)\n",
        "        # self.decode_conv5 = conv(64, 32, kernel_size)\n",
        "        self.decode_conv1 = nn.Sequential(\n",
        "            depthwise(1024, kernel_size),\n",
        "            pointwise(1024, 512))\n",
        "        self.decode_conv2 = nn.Sequential(\n",
        "            depthwise(512, kernel_size),\n",
        "            pointwise(512, 256))\n",
        "        self.decode_conv3 = nn.Sequential(\n",
        "            depthwise(512, kernel_size),\n",
        "            pointwise(512, 128))\n",
        "        self.decode_conv4 = nn.Sequential(\n",
        "            depthwise(256, kernel_size),\n",
        "            pointwise(256, 64))\n",
        "        self.decode_conv5 = nn.Sequential(\n",
        "            depthwise(128, kernel_size),\n",
        "            pointwise(128, 32))\n",
        "        self.decode_conv6 = pointwise(32, 1)\n",
        "        weights_init(self.decode_conv1)\n",
        "        weights_init(self.decode_conv2)\n",
        "        weights_init(self.decode_conv3)\n",
        "        weights_init(self.decode_conv4)\n",
        "        weights_init(self.decode_conv5)\n",
        "        weights_init(self.decode_conv6)\n",
        "\n",
        "    def forward(self, x):\n",
        "        # skip connections: dec4: enc1\n",
        "        # dec 3: enc2 or enc3\n",
        "        # dec 2: enc4 or enc5\n",
        "        for i in range(14):\n",
        "            layer = getattr(self, 'conv{}'.format(i))\n",
        "            x = layer(x)\n",
        "            # print(\"{}: {}\".format(i, x.size()))\n",
        "            if i==1:\n",
        "                x1 = x\n",
        "            elif i==3:\n",
        "                x2 = x\n",
        "            elif i==5:\n",
        "                x3 = x\n",
        "        for i in range(1,6):\n",
        "            layer = getattr(self, 'decode_conv{}'.format(i))\n",
        "            # print(\"{}a: {}\".format(i, x.size()))\n",
        "            x = layer(x)\n",
        "            # print(\"{}b: {}\".format(i, x.size()))\n",
        "            x = F.interpolate(x, scale_factor=2, mode='nearest')\n",
        "            if i==4:\n",
        "                x = torch.cat((x, x1), 1)\n",
        "            elif i==3:\n",
        "                x = torch.cat((x, x2), 1)\n",
        "            elif i==2:\n",
        "                x = torch.cat((x, x3), 1)\n",
        "            # print(\"{}c: {}\".format(i, x.size()))\n",
        "        x = self.decode_conv6(x)\n",
        "        # print('END', x.size())\n",
        "        return x\n",
        "\n",
        "class Result(object):\n",
        "    def __init__(self):\n",
        "        self.irmse, self.imae = 0, 0\n",
        "        self.mse, self.rmse, self.mae = 0, 0, 0\n",
        "        self.absrel, self.lg10 = 0, 0\n",
        "        self.delta1, self.delta2, self.delta3 = 0, 0, 0\n",
        "        self.data_time, self.gpu_time = 0, 0\n",
        "\n",
        "    def set_to_worst(self):\n",
        "        self.irmse, self.imae = np.inf, np.inf\n",
        "        self.mse, self.rmse, self.mae = np.inf, np.inf, np.inf\n",
        "        self.absrel, self.lg10 = np.inf, np.inf\n",
        "        self.delta1, self.delta2, self.delta3 = 0, 0, 0\n",
        "        self.data_time, self.gpu_time = 0, 0\n",
        "\n",
        "    def update(self, irmse, imae, mse, rmse, mae, absrel, lg10, delta1, delta2, delta3, gpu_time, data_time):\n",
        "        self.irmse, self.imae = irmse, imae\n",
        "        self.mse, self.rmse, self.mae = mse, rmse, mae\n",
        "        self.absrel, self.lg10 = absrel, lg10\n",
        "        self.delta1, self.delta2, self.delta3 = delta1, delta2, delta3\n",
        "        self.data_time, self.gpu_time = data_time, gpu_time\n",
        "\n",
        "    def evaluate(self, output, target):\n",
        "        valid_mask = ((target>0) + (output>0)) > 0\n",
        "\n",
        "        output = 1e3 * output[valid_mask]\n",
        "        target = 1e3 * target[valid_mask]\n",
        "        abs_diff = (output - target).abs()\n",
        "\n",
        "        self.mse = float((torch.pow(abs_diff, 2)).mean())\n",
        "        self.rmse = math.sqrt(self.mse)\n",
        "        self.mae = float(abs_diff.mean())\n",
        "        self.lg10 = float((log10(output) - log10(target)).abs().mean())\n",
        "        self.absrel = float((abs_diff / target).mean())\n",
        "\n",
        "        maxRatio = torch.max(output / target, target / output)\n",
        "        self.delta1 = float((maxRatio < 1.25).float().mean())\n",
        "        self.delta2 = float((maxRatio < 1.25 ** 2).float().mean())\n",
        "        self.delta3 = float((maxRatio < 1.25 ** 3).float().mean())\n",
        "        self.data_time = 0\n",
        "        self.gpu_time = 0\n",
        "\n",
        "        inv_output = 1 / output\n",
        "        inv_target = 1 / target\n",
        "        abs_inv_diff = (inv_output - inv_target).abs()\n",
        "        self.irmse = math.sqrt((torch.pow(abs_inv_diff, 2)).mean())\n",
        "        self.imae = float(abs_inv_diff.mean())\n",
        "\n"
      ],
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GnUKh-96MzPa"
      },
      "source": [
        "Loads Model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "n0F4X7nFH2EU"
      },
      "source": [
        "import torch\n",
        "checkpoint = torch.load(\"/content/drive/MyDrive/results/Dataset=nyudepthv2 -- Network=mobilenet -- Loss=BerHu -- lr=0.01 -- bs=8/model_best.pth.tar\")\n",
        "model = checkpoint['model']"
      ],
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "awdoBhwFM1zl"
      },
      "source": [
        "Produce Prediction"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0vlwbM65IPBH",
        "outputId": "2ba8333a-9cca-490d-9f82-ce8893dd54ef"
      },
      "source": [
        "from PIL import Image\n",
        "from torchvision.transforms import ToTensor\n",
        "from torch.autograd import Variable\n",
        "import torch.nn.functional as F\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "img_path=\"/content/3d-photo-inpainting/image/ReDWeb_Sample.jpg\"\n",
        "image = Image.open(img_path)\n",
        "image = ToTensor()(image).unsqueeze(0) # unsqueeze to add artificial first dimension\n",
        "image = Variable(image)\n",
        "image = image.cuda()\n",
        "print(image.shape)\n",
        "\n",
        "pred = model(image.to(dtype=torch.float))\n",
        "pred_c = pred.squeeze(0).squeeze(0).cpu().detach().numpy()"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "torch.Size([1, 3, 224, 224])\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 287
        },
        "id": "Ua8wExK0IcMW",
        "outputId": "6b2f71b8-61bc-433a-a393-da95e603e537"
      },
      "source": [
        "plt.imshow(pred_c)"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7ff28464bcd0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 16
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vtVfYHCINALX"
      },
      "source": [
        "Save depth as npy and png format"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "id": "Z0z1JlePKVZ7",
        "outputId": "38d67c8c-3091-4974-f5d6-f75500f5b042"
      },
      "source": [
        "import numpy as np\n",
        "np.save(\"/content/3d-photo-inpainting/depth/ReDWeb_Sample.npy\",pred_c)\n",
        "plt.savefig(\"/content/3d-photo-inpainting/depth/ReDWeb_Sample.png\")"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 0 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "U6LV9lm2LcCI"
      },
      "source": [
        "- **Execute the 3D Photo Inpainting**\n",
        "  - Note: The 3D photo generation process usually takes about 2-3 minutes or more depending on the available computing resources."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DoqjT7dkKuyO",
        "outputId": "73f1dc54-f0ed-4362-b09b-25b8ae20a2e4"
      },
      "source": [
        "!python main.py --config argument.yml"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "running on device 0\n",
            "\r  0% 0/1 [00:00<?, ?it/s]Current Source ==>  ReDWeb_Sample\n",
            "Running depth extraction at 1616689049.1634347\n",
            "Start Running 3D_Photo ...\n",
            "Loading edge model at 1616689061.2317262\n",
            "Loading depth model at 1616689063.5167317\n",
            "Loading rgb model at 1616689064.4793072\n",
            "Writing depth ply (and basically doing everything) at 1616689065.4013197\n",
            "Writing mesh file mesh/ReDWeb_Sample.ply ...\n",
            "Making video at 1616689246.0316343\n",
            "fov: 53.13010235415598\n",
            "Moviepy - Building video video/ReDWeb_Sample_dolly-zoom-in.mp4.\n",
            "Moviepy - Writing video video/ReDWeb_Sample_dolly-zoom-in.mp4\n",
            "\n",
            "  0% 0/1 [03:59<?, ?it/s]\n",
            "t:   0% 0/241 [00:00<?, ?it/s, now=None]\u001b[A\n",
            "t:   1% 2/241 [00:00<00:15, 15.59it/s, now=None]\u001b[A\n",
            "t:   1% 3/241 [00:00<00:20, 11.79it/s, now=None]\u001b[A\n",
            "t:   6% 14/241 [00:00<00:14, 16.10it/s, now=None]\u001b[A\n",
            "t:  10% 25/241 [00:00<00:10, 21.55it/s, now=None]\u001b[A\n",
            "t:  15% 36/241 [00:00<00:07, 28.22it/s, now=None]\u001b[A\n",
            "t:  19% 46/241 [00:00<00:05, 35.88it/s, now=None]\u001b[A\n",
            "t:  22% 54/241 [00:01<00:06, 28.02it/s, now=None]\u001b[A\n",
            "t:  25% 61/241 [00:01<00:06, 28.06it/s, now=None]\u001b[A\n",
            "t:  28% 67/241 [00:01<00:06, 26.11it/s, now=None]\u001b[A\n",
            "t:  30% 72/241 [00:01<00:06, 24.81it/s, now=None]\u001b[A\n",
            "t:  32% 76/241 [00:02<00:07, 23.19it/s, now=None]\u001b[A\n",
            "t:  33% 80/241 [00:02<00:07, 22.76it/s, now=None]\u001b[A\n",
            "t:  34% 83/241 [00:02<00:06, 22.89it/s, now=None]\u001b[A\n",
            "t:  36% 86/241 [00:02<00:07, 20.56it/s, now=None]\u001b[A\n",
            "t:  37% 89/241 [00:02<00:07, 19.40it/s, now=None]\u001b[A\n",
            "t:  38% 92/241 [00:02<00:07, 20.60it/s, now=None]\u001b[A\n",
            "t:  39% 95/241 [00:02<00:06, 21.86it/s, now=None]\u001b[A\n",
            "t:  41% 98/241 [00:03<00:07, 19.73it/s, now=None]\u001b[A\n",
            "t:  42% 101/241 [00:03<00:07, 19.92it/s, now=None]\u001b[A\n",
            "t:  44% 105/241 [00:03<00:06, 20.09it/s, now=None]\u001b[A\n",
            "t:  45% 108/241 [00:03<00:06, 20.45it/s, now=None]\u001b[A\n",
            "t:  46% 111/241 [00:03<00:06, 20.64it/s, now=None]\u001b[A\n",
            "t:  47% 114/241 [00:03<00:06, 19.79it/s, now=None]\u001b[A\n",
            "t:  49% 117/241 [00:04<00:06, 19.51it/s, now=None]\u001b[A\n",
            "t:  50% 120/241 [00:04<00:05, 21.46it/s, now=None]\u001b[A\n",
            "t:  51% 123/241 [00:04<00:05, 21.80it/s, now=None]\u001b[A\n",
            "t:  52% 126/241 [00:04<00:05, 21.05it/s, now=None]\u001b[A\n",
            "t:  54% 129/241 [00:04<00:05, 20.53it/s, now=None]\u001b[A\n",
            "t:  55% 133/241 [00:04<00:05, 20.20it/s, now=None]\u001b[A\n",
            "t:  57% 137/241 [00:05<00:05, 20.68it/s, now=None]\u001b[A\n",
            "t:  59% 141/241 [00:05<00:04, 20.66it/s, now=None]\u001b[A\n",
            "t:  60% 145/241 [00:05<00:04, 20.74it/s, now=None]\u001b[A\n",
            "t:  61% 148/241 [00:05<00:04, 22.84it/s, now=None]\u001b[A\n",
            "t:  63% 151/241 [00:05<00:04, 21.86it/s, now=None]\u001b[A\n",
            "t:  64% 154/241 [00:05<00:04, 21.08it/s, now=None]\u001b[A\n",
            "t:  65% 157/241 [00:05<00:04, 20.95it/s, now=None]\u001b[A\n",
            "t:  67% 161/241 [00:06<00:03, 20.58it/s, now=None]\u001b[A\n",
            "t:  68% 165/241 [00:06<00:03, 22.49it/s, now=None]\u001b[A\n",
            "t:  70% 169/241 [00:06<00:02, 25.49it/s, now=None]\u001b[A\n",
            "t:  72% 174/241 [00:06<00:02, 28.60it/s, now=None]\u001b[A\n",
            "t:  74% 178/241 [00:06<00:02, 28.95it/s, now=None]\u001b[A\n",
            "t:  76% 182/241 [00:06<00:02, 28.28it/s, now=None]\u001b[A\n",
            "t:  77% 185/241 [00:06<00:02, 24.67it/s, now=None]\u001b[A\n",
            "t:  78% 188/241 [00:07<00:02, 22.82it/s, now=None]\u001b[A\n",
            "t:  79% 191/241 [00:07<00:02, 21.91it/s, now=None]\u001b[A\n",
            "t:  80% 194/241 [00:07<00:02, 22.70it/s, now=None]\u001b[A\n",
            "t:  82% 197/241 [00:07<00:01, 23.28it/s, now=None]\u001b[A\n",
            "t:  83% 200/241 [00:07<00:01, 21.32it/s, now=None]\u001b[A\n",
            "t:  84% 203/241 [00:07<00:01, 19.56it/s, now=None]\u001b[A\n",
            "t:  85% 206/241 [00:08<00:01, 19.91it/s, now=None]\u001b[A\n",
            "t:  87% 210/241 [00:08<00:01, 22.07it/s, now=None]\u001b[A\n",
            "t:  88% 213/241 [00:08<00:01, 21.51it/s, now=None]\u001b[A\n",
            "t:  90% 216/241 [00:08<00:01, 20.91it/s, now=None]\u001b[A\n",
            "t:  91% 219/241 [00:08<00:01, 19.58it/s, now=None]\u001b[A\n",
            "t:  93% 223/241 [00:08<00:00, 20.12it/s, now=None]\u001b[A\n",
            "t:  94% 227/241 [00:09<00:00, 19.60it/s, now=None]\u001b[A\n",
            "t:  95% 229/241 [00:09<00:00, 18.87it/s, now=None]\u001b[A\n",
            "t:  96% 232/241 [00:09<00:00, 21.13it/s, now=None]\u001b[A\n",
            "t:  98% 235/241 [00:09<00:00, 19.46it/s, now=None]\u001b[A\n",
            "t:  99% 239/241 [00:09<00:00, 20.47it/s, now=None]\u001b[A\n",
            "Moviepy - Done !\n",
            "Moviepy - video ready video/ReDWeb_Sample_dolly-zoom-in.mp4\n",
            "Moviepy - Building video video/ReDWeb_Sample_zoom-in.mp4.\n",
            "Moviepy - Writing video video/ReDWeb_Sample_zoom-in.mp4\n",
            "\n",
            "  0% 0/1 [04:48<?, ?it/s]\n",
            "t:   0% 0/241 [00:00<?, ?it/s, now=None]\u001b[A\n",
            "t:   4% 10/241 [00:00<00:02, 94.97it/s, now=None]\u001b[A\n",
            "t:   9% 21/241 [00:00<00:02, 97.42it/s, now=None]\u001b[A\n",
            "t:  13% 32/241 [00:00<00:02, 99.98it/s, now=None]\u001b[A\n",
            "t:  18% 44/241 [00:00<00:01, 102.54it/s, now=None]\u001b[A\n",
            "t:  22% 53/241 [00:00<00:04, 41.93it/s, now=None] \u001b[A\n",
            "t:  25% 60/241 [00:01<00:05, 32.37it/s, now=None]\u001b[A\n",
            "t:  27% 65/241 [00:01<00:05, 30.11it/s, now=None]\u001b[A\n",
            "t:  29% 70/241 [00:01<00:06, 26.58it/s, now=None]\u001b[A\n",
            "t:  31% 74/241 [00:01<00:06, 24.53it/s, now=None]\u001b[A\n",
            "t:  32% 78/241 [00:02<00:07, 22.44it/s, now=None]\u001b[A\n",
            "t:  34% 81/241 [00:02<00:07, 21.17it/s, now=None]\u001b[A\n",
            "t:  35% 84/241 [00:02<00:07, 21.24it/s, now=None]\u001b[A\n",
            "t:  36% 87/241 [00:02<00:07, 20.69it/s, now=None]\u001b[A\n",
            "t:  37% 90/241 [00:02<00:07, 20.53it/s, now=None]\u001b[A\n",
            "t:  39% 93/241 [00:02<00:06, 21.20it/s, now=None]\u001b[A\n",
            "t:  40% 96/241 [00:02<00:06, 21.78it/s, now=None]\u001b[A\n",
            "t:  41% 99/241 [00:03<00:06, 20.39it/s, now=None]\u001b[A\n",
            "t:  42% 102/241 [00:03<00:06, 21.37it/s, now=None]\u001b[A\n",
            "t:  44% 105/241 [00:03<00:06, 21.43it/s, now=None]\u001b[A\n",
            "t:  45% 108/241 [00:03<00:06, 20.05it/s, now=None]\u001b[A\n",
            "t:  46% 111/241 [00:03<00:06, 19.12it/s, now=None]\u001b[A\n",
            "t:  48% 115/241 [00:03<00:06, 19.44it/s, now=None]\u001b[A\n",
            "t:  49% 118/241 [00:04<00:05, 20.81it/s, now=None]\u001b[A\n",
            "t:  50% 121/241 [00:04<00:05, 21.10it/s, now=None]\u001b[A\n",
            "t:  51% 124/241 [00:04<00:05, 19.94it/s, now=None]\u001b[A\n",
            "t:  53% 127/241 [00:04<00:05, 19.44it/s, now=None]\u001b[A\n",
            "t:  54% 130/241 [00:04<00:05, 21.16it/s, now=None]\u001b[A\n",
            "t:  55% 133/241 [00:04<00:05, 19.62it/s, now=None]\u001b[A\n",
            "t:  56% 136/241 [00:04<00:05, 19.78it/s, now=None]\u001b[A\n",
            "t:  58% 139/241 [00:05<00:05, 19.35it/s, now=None]\u001b[A\n",
            "t:  59% 142/241 [00:05<00:04, 21.01it/s, now=None]\u001b[A\n",
            "t:  60% 145/241 [00:05<00:04, 21.38it/s, now=None]\u001b[A\n",
            "t:  61% 148/241 [00:05<00:04, 19.96it/s, now=None]\u001b[A\n",
            "t:  63% 151/241 [00:05<00:04, 19.99it/s, now=None]\u001b[A\n",
            "t:  64% 154/241 [00:05<00:03, 21.97it/s, now=None]\u001b[A\n",
            "t:  65% 157/241 [00:05<00:03, 21.08it/s, now=None]\u001b[A\n",
            "t:  66% 160/241 [00:06<00:04, 19.35it/s, now=None]\u001b[A\n",
            "t:  68% 163/241 [00:06<00:04, 19.25it/s, now=None]\u001b[A\n",
            "t:  69% 167/241 [00:06<00:03, 18.93it/s, now=None]\u001b[A\n",
            "t:  71% 171/241 [00:06<00:03, 19.63it/s, now=None]\u001b[A\n",
            "t:  73% 175/241 [00:06<00:03, 20.02it/s, now=None]\u001b[A\n",
            "t:  74% 179/241 [00:07<00:03, 20.47it/s, now=None]\u001b[A\n",
            "t:  76% 182/241 [00:07<00:02, 20.74it/s, now=None]\u001b[A\n",
            "t:  77% 185/241 [00:07<00:02, 21.51it/s, now=None]\u001b[A\n",
            "t:  78% 188/241 [00:07<00:02, 20.53it/s, now=None]\u001b[A\n",
            "t:  79% 191/241 [00:07<00:02, 19.70it/s, now=None]\u001b[A\n",
            "t:  80% 194/241 [00:07<00:02, 19.90it/s, now=None]\u001b[A\n",
            "t:  82% 197/241 [00:07<00:02, 21.76it/s, now=None]\u001b[A\n",
            "t:  83% 200/241 [00:08<00:01, 20.62it/s, now=None]\u001b[A\n",
            "t:  84% 203/241 [00:08<00:02, 18.65it/s, now=None]\u001b[A\n",
            "t:  86% 207/241 [00:08<00:01, 19.20it/s, now=None]\u001b[A\n",
            "t:  88% 211/241 [00:08<00:01, 19.63it/s, now=None]\u001b[A\n",
            "t:  89% 214/241 [00:08<00:01, 19.67it/s, now=None]\u001b[A\n",
            "t:  90% 217/241 [00:08<00:01, 21.76it/s, now=None]\u001b[A\n",
            "t:  91% 220/241 [00:09<00:01, 20.74it/s, now=None]\u001b[A\n",
            "t:  93% 223/241 [00:09<00:00, 20.07it/s, now=None]\u001b[A\n",
            "t:  94% 227/241 [00:09<00:00, 20.18it/s, now=None]\u001b[A\n",
            "t:  95% 230/241 [00:09<00:00, 21.70it/s, now=None]\u001b[A\n",
            "t:  97% 233/241 [00:09<00:00, 22.19it/s, now=None]\u001b[A\n",
            "t:  98% 236/241 [00:09<00:00, 20.54it/s, now=None]\u001b[A\n",
            "t:  99% 239/241 [00:10<00:00, 19.63it/s, now=None]\u001b[A\n",
            "Moviepy - Done !\n",
            "Moviepy - video ready video/ReDWeb_Sample_zoom-in.mp4\n",
            "Moviepy - Building video video/ReDWeb_Sample_circle.mp4.\n",
            "Moviepy - Writing video video/ReDWeb_Sample_circle.mp4\n",
            "\n",
            "  0% 0/1 [05:39<?, ?it/s]\n",
            "t:   0% 0/241 [00:00<?, ?it/s, now=None]\u001b[A\n",
            "t:   4% 10/241 [00:00<00:02, 97.65it/s, now=None]\u001b[A\n",
            "t:   9% 22/241 [00:00<00:02, 102.87it/s, now=None]\u001b[A\n",
            "t:  13% 31/241 [00:00<00:02, 97.46it/s, now=None] \u001b[A\n",
            "t:  16% 38/241 [00:00<00:02, 84.71it/s, now=None]\u001b[A\n",
            "t:  20% 47/241 [00:00<00:02, 84.88it/s, now=None]\u001b[A\n",
            "t:  23% 55/241 [00:01<00:06, 29.50it/s, now=None]\u001b[A\n",
            "t:  25% 61/241 [00:01<00:06, 26.43it/s, now=None]\u001b[A\n",
            "t:  27% 66/241 [00:01<00:07, 24.35it/s, now=None]\u001b[A\n",
            "t:  29% 70/241 [00:01<00:08, 21.08it/s, now=None]\u001b[A\n",
            "t:  31% 74/241 [00:02<00:08, 20.14it/s, now=None]\u001b[A\n",
            "t:  32% 77/241 [00:02<00:08, 19.56it/s, now=None]\u001b[A\n",
            "t:  33% 80/241 [00:02<00:08, 17.89it/s, now=None]\u001b[A\n",
            "t:  34% 83/241 [00:02<00:09, 17.10it/s, now=None]\u001b[A\n",
            "t:  35% 85/241 [00:02<00:09, 16.31it/s, now=None]\u001b[A\n",
            "t:  37% 88/241 [00:03<00:08, 18.10it/s, now=None]\u001b[A\n",
            "t:  38% 91/241 [00:03<00:08, 18.62it/s, now=None]\u001b[A\n",
            "t:  39% 94/241 [00:03<00:07, 20.11it/s, now=None]\u001b[A\n",
            "t:  40% 97/241 [00:03<00:07, 19.28it/s, now=None]\u001b[A\n",
            "t:  41% 100/241 [00:03<00:07, 18.02it/s, now=None]\u001b[A\n",
            "t:  42% 102/241 [00:03<00:07, 18.21it/s, now=None]\u001b[A\n",
            "t:  43% 104/241 [00:03<00:07, 18.56it/s, now=None]\u001b[A\n",
            "t:  44% 106/241 [00:03<00:07, 18.81it/s, now=None]\u001b[A\n",
            "t:  45% 108/241 [00:04<00:07, 18.28it/s, now=None]\u001b[A\n",
            "t:  46% 110/241 [00:04<00:07, 18.63it/s, now=None]\u001b[A\n",
            "t:  46% 112/241 [00:04<00:07, 18.35it/s, now=None]\u001b[A\n",
            "t:  48% 115/241 [00:04<00:07, 17.40it/s, now=None]\u001b[A\n",
            "t:  49% 118/241 [00:04<00:06, 18.73it/s, now=None]\u001b[A\n",
            "t:  50% 120/241 [00:04<00:06, 18.50it/s, now=None]\u001b[A\n",
            "t:  51% 123/241 [00:04<00:06, 18.33it/s, now=None]\u001b[A\n",
            "t:  52% 126/241 [00:05<00:06, 19.05it/s, now=None]\u001b[A\n",
            "t:  53% 128/241 [00:05<00:06, 18.40it/s, now=None]\u001b[A\n",
            "t:  54% 130/241 [00:05<00:06, 18.08it/s, now=None]\u001b[A\n",
            "t:  55% 132/241 [00:05<00:05, 18.28it/s, now=None]\u001b[A\n",
            "t:  56% 134/241 [00:05<00:06, 17.45it/s, now=None]\u001b[A\n",
            "t:  56% 136/241 [00:05<00:05, 17.78it/s, now=None]\u001b[A\n",
            "t:  58% 139/241 [00:05<00:05, 17.42it/s, now=None]\u001b[A\n",
            "t:  59% 142/241 [00:05<00:05, 18.65it/s, now=None]\u001b[A\n",
            "t:  60% 144/241 [00:06<00:05, 18.31it/s, now=None]\u001b[A\n",
            "t:  61% 147/241 [00:06<00:05, 17.85it/s, now=None]\u001b[A\n",
            "t:  62% 150/241 [00:06<00:04, 19.62it/s, now=None]\u001b[A\n",
            "t:  63% 153/241 [00:06<00:04, 19.00it/s, now=None]\u001b[A\n",
            "t:  64% 155/241 [00:06<00:05, 16.67it/s, now=None]\u001b[A\n",
            "t:  66% 158/241 [00:06<00:04, 18.63it/s, now=None]\u001b[A\n",
            "t:  66% 160/241 [00:06<00:04, 18.48it/s, now=None]\u001b[A\n",
            "t:  68% 163/241 [00:07<00:04, 17.26it/s, now=None]\u001b[A\n",
            "t:  69% 166/241 [00:07<00:04, 18.43it/s, now=None]\u001b[A\n",
            "t:  70% 169/241 [00:07<00:03, 19.08it/s, now=None]\u001b[A\n",
            "t:  71% 171/241 [00:07<00:04, 17.24it/s, now=None]\u001b[A\n",
            "t:  72% 174/241 [00:07<00:03, 19.37it/s, now=None]\u001b[A\n",
            "t:  73% 177/241 [00:07<00:03, 19.79it/s, now=None]\u001b[A\n",
            "t:  75% 180/241 [00:07<00:03, 18.47it/s, now=None]\u001b[A\n",
            "t:  76% 182/241 [00:08<00:03, 17.94it/s, now=None]\u001b[A\n",
            "t:  77% 185/241 [00:08<00:02, 19.60it/s, now=None]\u001b[A\n",
            "t:  78% 188/241 [00:08<00:02, 18.81it/s, now=None]\u001b[A\n",
            "t:  79% 190/241 [00:08<00:02, 18.59it/s, now=None]\u001b[A\n",
            "t:  80% 192/241 [00:08<00:02, 17.37it/s, now=None]\u001b[A\n",
            "t:  81% 195/241 [00:08<00:02, 17.53it/s, now=None]\u001b[A\n",
            "t:  82% 198/241 [00:08<00:02, 19.87it/s, now=None]\u001b[A\n",
            "t:  83% 201/241 [00:09<00:02, 18.54it/s, now=None]\u001b[A\n",
            "t:  84% 203/241 [00:09<00:02, 17.44it/s, now=None]\u001b[A\n",
            "t:  85% 206/241 [00:09<00:01, 18.62it/s, now=None]\u001b[A\n",
            "t:  86% 208/241 [00:09<00:01, 17.45it/s, now=None]\u001b[A\n",
            "t:  88% 211/241 [00:09<00:01, 16.95it/s, now=None]\u001b[A\n",
            "t:  89% 214/241 [00:09<00:01, 17.59it/s, now=None]\u001b[A\n",
            "t:  90% 217/241 [00:09<00:01, 19.00it/s, now=None]\u001b[A\n",
            "t:  91% 219/241 [00:10<00:01, 16.50it/s, now=None]\u001b[A\n",
            "t:  92% 222/241 [00:10<00:01, 17.44it/s, now=None]\u001b[A\n",
            "t:  93% 224/241 [00:10<00:00, 17.26it/s, now=None]\u001b[A\n",
            "t:  94% 227/241 [00:10<00:00, 16.91it/s, now=None]\u001b[A\n",
            "t:  95% 230/241 [00:10<00:00, 19.34it/s, now=None]\u001b[A\n",
            "t:  97% 233/241 [00:10<00:00, 18.11it/s, now=None]\u001b[A\n",
            "t:  98% 235/241 [00:11<00:00, 15.82it/s, now=None]\u001b[A\n",
            "t:  99% 238/241 [00:11<00:00, 17.20it/s, now=None]\u001b[A\n",
            "t: 100% 240/241 [00:11<00:00, 16.85it/s, now=None]\u001b[A\n",
            "Moviepy - Done !\n",
            "Moviepy - video ready video/ReDWeb_Sample_circle.mp4\n",
            "Moviepy - Building video video/ReDWeb_Sample_swing.mp4.\n",
            "Moviepy - Writing video video/ReDWeb_Sample_swing.mp4\n",
            "\n",
            "  0% 0/1 [06:31<?, ?it/s]\n",
            "t:   0% 0/241 [00:00<?, ?it/s, now=None]\u001b[A\n",
            "t:   1% 3/241 [00:00<00:11, 21.54it/s, now=None]\u001b[A\n",
            "t:   5% 13/241 [00:00<00:08, 28.04it/s, now=None]\u001b[A\n",
            "t:  10% 25/241 [00:00<00:05, 36.30it/s, now=None]\u001b[A\n",
            "t:  15% 37/241 [00:00<00:04, 45.52it/s, now=None]\u001b[A\n",
            "t:  20% 47/241 [00:00<00:03, 53.41it/s, now=None]\u001b[A\n",
            "t:  23% 55/241 [00:01<00:06, 28.87it/s, now=None]\u001b[A\n",
            "t:  25% 61/241 [00:01<00:07, 24.46it/s, now=None]\u001b[A\n",
            "t:  27% 66/241 [00:01<00:07, 22.24it/s, now=None]\u001b[A\n",
            "t:  29% 70/241 [00:01<00:07, 22.72it/s, now=None]\u001b[A\n",
            "t:  31% 74/241 [00:02<00:07, 21.25it/s, now=None]\u001b[A\n",
            "t:  32% 77/241 [00:02<00:08, 18.64it/s, now=None]\u001b[A\n",
            "t:  33% 80/241 [00:02<00:08, 18.29it/s, now=None]\u001b[A\n",
            "t:  35% 84/241 [00:02<00:08, 18.61it/s, now=None]\u001b[A\n",
            "t:  37% 88/241 [00:02<00:08, 18.72it/s, now=None]\u001b[A\n",
            "t:  38% 92/241 [00:03<00:08, 18.60it/s, now=None]\u001b[A\n",
            "t:  40% 96/241 [00:03<00:07, 18.57it/s, now=None]\u001b[A\n",
            "t:  41% 99/241 [00:03<00:07, 19.49it/s, now=None]\u001b[A\n",
            "t:  42% 102/241 [00:03<00:06, 19.95it/s, now=None]\u001b[A\n",
            "t:  44% 105/241 [00:03<00:07, 18.15it/s, now=None]\u001b[A\n",
            "t:  45% 108/241 [00:04<00:07, 17.35it/s, now=None]\u001b[A\n",
            "t:  46% 110/241 [00:04<00:07, 17.53it/s, now=None]\u001b[A\n",
            "t:  47% 113/241 [00:04<00:07, 18.26it/s, now=None]\u001b[A\n",
            "t:  48% 116/241 [00:04<00:07, 17.26it/s, now=None]\u001b[A\n",
            "t:  50% 120/241 [00:04<00:06, 17.79it/s, now=None]\u001b[A\n",
            "t:  51% 124/241 [00:04<00:06, 18.12it/s, now=None]\u001b[A\n",
            "t:  53% 128/241 [00:05<00:06, 18.36it/s, now=None]\u001b[A\n",
            "t:  54% 131/241 [00:05<00:05, 19.73it/s, now=None]\u001b[A\n",
            "t:  56% 134/241 [00:05<00:05, 20.78it/s, now=None]\u001b[A\n",
            "t:  57% 137/241 [00:05<00:05, 18.61it/s, now=None]\u001b[A\n",
            "t:  58% 140/241 [00:05<00:05, 17.87it/s, now=None]\u001b[A\n",
            "t:  60% 144/241 [00:05<00:05, 18.19it/s, now=None]\u001b[A\n",
            "t:  61% 147/241 [00:06<00:04, 19.19it/s, now=None]\u001b[A\n",
            "t:  62% 149/241 [00:06<00:04, 18.79it/s, now=None]\u001b[A\n",
            "t:  63% 152/241 [00:06<00:04, 18.37it/s, now=None]\u001b[A\n",
            "t:  64% 155/241 [00:06<00:04, 19.67it/s, now=None]\u001b[A\n",
            "t:  66% 158/241 [00:06<00:03, 21.51it/s, now=None]\u001b[A\n",
            "t:  67% 161/241 [00:06<00:04, 19.91it/s, now=None]\u001b[A\n",
            "t:  68% 164/241 [00:06<00:03, 19.29it/s, now=None]\u001b[A\n",
            "t:  69% 167/241 [00:07<00:03, 21.39it/s, now=None]\u001b[A\n",
            "t:  71% 171/241 [00:07<00:03, 21.70it/s, now=None]\u001b[A\n",
            "t:  72% 174/241 [00:07<00:03, 22.19it/s, now=None]\u001b[A\n",
            "t:  73% 177/241 [00:07<00:02, 22.57it/s, now=None]\u001b[A\n",
            "t:  75% 180/241 [00:07<00:02, 21.53it/s, now=None]\u001b[A\n",
            "t:  76% 183/241 [00:07<00:02, 22.10it/s, now=None]\u001b[A\n",
            "t:  77% 186/241 [00:07<00:02, 21.35it/s, now=None]\u001b[A\n",
            "t:  78% 189/241 [00:08<00:02, 20.27it/s, now=None]\u001b[A\n",
            "t:  80% 192/241 [00:08<00:02, 19.33it/s, now=None]\u001b[A\n",
            "t:  80% 194/241 [00:08<00:02, 18.30it/s, now=None]\u001b[A\n",
            "t:  82% 197/241 [00:08<00:02, 20.15it/s, now=None]\u001b[A\n",
            "t:  83% 200/241 [00:08<00:02, 19.01it/s, now=None]\u001b[A\n",
            "t:  84% 203/241 [00:08<00:01, 19.48it/s, now=None]\u001b[A\n",
            "t:  85% 206/241 [00:08<00:01, 21.14it/s, now=None]\u001b[A\n",
            "t:  87% 209/241 [00:09<00:01, 18.85it/s, now=None]\u001b[A\n",
            "t:  88% 211/241 [00:09<00:01, 18.17it/s, now=None]\u001b[A\n",
            "t:  89% 214/241 [00:09<00:01, 18.81it/s, now=None]\u001b[A\n",
            "t:  90% 216/241 [00:09<00:01, 18.09it/s, now=None]\u001b[A\n",
            "t:  91% 220/241 [00:09<00:01, 18.54it/s, now=None]\u001b[A\n",
            "t:  93% 224/241 [00:09<00:00, 18.81it/s, now=None]\u001b[A\n",
            "t:  94% 226/241 [00:10<00:00, 18.30it/s, now=None]\u001b[A\n",
            "t:  95% 229/241 [00:10<00:00, 18.91it/s, now=None]\u001b[A\n",
            "t:  96% 232/241 [00:10<00:00, 18.58it/s, now=None]\u001b[A\n",
            "t:  98% 236/241 [00:10<00:00, 18.71it/s, now=None]\u001b[A\n",
            "t: 100% 240/241 [00:10<00:00, 18.75it/s, now=None]\u001b[A\n",
            "Moviepy - Done !\n",
            "Moviepy - video ready video/ReDWeb_Sample_swing.mp4\n",
            "100% 1/1 [06:44<00:00, 404.74s/it]\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}