CNN_Denoiser_AE.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "CNN Denoiser OG.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "XNYJgzG1ZPul"
      },
      "source": [
        "import torch\n",
        "import torchvision\n",
        "from torchvision import datasets,transforms\n"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rnEtPM6CZbNr",
        "outputId": "146a1ec3-79c1-4701-8d05-7b5d4a9f6711"
      },
      "source": [
        "train = datasets.MNIST('',train = True, download = True, transform = transforms.Compose([transforms.ToTensor()]))\n",
        "trainset = torch.utils.data.DataLoader(train,batch_size = 32,shuffle = True)\n",
        "test = datasets.MNIST('',train = False, download = True, transform = transforms.Compose([transforms.ToTensor()]))\n",
        "testset = torch.utils.data.DataLoader(test,batch_size = 32,shuffle = True)\n",
        "print(train.data.size())"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "torch.Size([60000, 28, 28])\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/torchvision/datasets/mnist.py:498: UserWarning: The given NumPy array is not writeable, and PyTorch does not support non-writeable tensors. This means you can write to the underlying (supposedly non-writeable) NumPy array using the tensor. You may want to copy the array to protect its data or make it writeable before converting it to a tensor. This type of warning will be suppressed for the rest of this program. (Triggered internally at  /pytorch/torch/csrc/utils/tensor_numpy.cpp:180.)\n",
            "  return torch.from_numpy(parsed.astype(m[2], copy=False)).view(*s)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_kEvSl7cbmSd"
      },
      "source": [
        "import torch.nn as nn\n",
        "import torch.nn.functional as F"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "p13p41nZl8Ek",
        "outputId": "54719372-5518-4b2f-9871-3e2a4636573d"
      },
      "source": [
        "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n",
        "print(device)"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "cuda\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PiXAiz8Dby7n",
        "outputId": "569c8e3f-ceb9-4ffa-d37a-41f947818f4c"
      },
      "source": [
        "class Model(nn.Module):\n",
        "  def __init__(self):\n",
        "    super().__init__()\n",
        "    #encoding layers\n",
        "    self.conv1 = nn.Conv2d(1,32,3,padding=1,stride=2) # 28x28 --> 14x14\n",
        "    self.conv2 = nn.Conv2d(32,64,3,padding=1,stride=2) # 14x14 --> 7x7\n",
        "    self.conv3 = nn.Conv2d(64,128,5) # 7x7 --> 3x3\n",
        "\n",
        "\n",
        "    #decoding layers\n",
        "    self.tconv1 = nn.ConvTranspose2d(128,64,5) # 3x3 --> 7x7\n",
        "    self.tconv2 = nn.ConvTranspose2d(64,32,3,stride=2,padding=1,output_padding=1) # 7x7 --> 14x14\n",
        "    self.tconv3 = nn.ConvTranspose2d(32,1,3,stride=2,padding=1,output_padding=1) # 14x14 --> 28x28\n",
        "\n",
        "  \n",
        "  #forward prop function \n",
        "  def forward(self,x):\n",
        "    x = F.relu(self.conv1(x))\n",
        "    # print(x.size())\n",
        "    x = F.relu(self.conv2(x))\n",
        "    # print(x.size())\n",
        "    x = F.relu(self.conv3(x))\n",
        "    # print(x.size())\n",
        "\n",
        "    x = F.relu(self.tconv1(x))\n",
        "    # print(x.size())\n",
        "    x = F.relu(self.tconv2(x))\n",
        "    # print(x.size())\n",
        "    x = F.sigmoid(self.tconv3(x)) #final layer is applied sigmoid activation\n",
        "    # print(x.size())\n",
        "\n",
        "    return x\n",
        "\n",
        "model = Model()\n",
        "print(model)\n"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model(\n",
            "  (conv1): Conv2d(1, 32, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))\n",
            "  (conv2): Conv2d(32, 64, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))\n",
            "  (conv3): Conv2d(64, 128, kernel_size=(5, 5), stride=(1, 1))\n",
            "  (tconv1): ConvTranspose2d(128, 64, kernel_size=(5, 5), stride=(1, 1))\n",
            "  (tconv2): ConvTranspose2d(64, 32, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), output_padding=(1, 1))\n",
            "  (tconv3): ConvTranspose2d(32, 1, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), output_padding=(1, 1))\n",
            ")\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_nxAfalAm5R3",
        "outputId": "f876fa04-4d77-4455-9270-84c7223c3b2c"
      },
      "source": [
        "model.to(device) #shifting model to cuda"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Model(\n",
              "  (conv1): Conv2d(1, 32, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))\n",
              "  (conv2): Conv2d(32, 64, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))\n",
              "  (conv3): Conv2d(64, 128, kernel_size=(5, 5), stride=(1, 1))\n",
              "  (tconv1): ConvTranspose2d(128, 64, kernel_size=(5, 5), stride=(1, 1))\n",
              "  (tconv2): ConvTranspose2d(64, 32, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), output_padding=(1, 1))\n",
              "  (tconv3): ConvTranspose2d(32, 1, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), output_padding=(1, 1))\n",
              ")"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LpUthJvwlWJy"
      },
      "source": [
        "import torch.optim as optim\n",
        "loss_function = nn.MSELoss() #mean-squared error loss_function\n",
        "optimizer = optim.Adam(model.parameters(),lr=0.001)"
      ],
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "K3z67Hfxlmkm"
      },
      "source": [
        "#function to add noise to input images\n",
        "def add_noise(inputs,noise_factor=0.3):\n",
        "  noisy = inputs + (torch.randn_like(inputs) * noise_factor)\n",
        "  noisy = torch.clip(noisy,0.,1.)\n",
        "  return noisy"
      ],
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "jn_4O2lRmA5U",
        "outputId": "b7db980b-18f5-4bb9-e229-d5efd6db2dcf"
      },
      "source": [
        "losses = [] #list containing average loss of each epoch\n",
        "running_loss = 0\n",
        "l = len(trainset)\n",
        "\n",
        "#training loop\n",
        "for epoch in range(10):\n",
        "  for t in trainset:\n",
        "    images,_ = t\n",
        "    noisy_images = add_noise(images)\n",
        "    images = images.to(device)\n",
        "    noisy_images = noisy_images.to(device) #shifting noisy images to cuda\n",
        "    \n",
        "    model.zero_grad()\n",
        "    output = model(noisy_images)\n",
        "    \n",
        "    loss = loss_function(output,images) #loss is calculated between model output and og dataset images\n",
        "    loss.backward()\n",
        "    optimizer.step()\n",
        "\n",
        "    running_loss+=loss.item() \n",
        "  losses.append(running_loss/l)\n",
        "  running_loss = 0\n",
        "  print(loss) #prints loss of the last batch of the epoch"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/torch/nn/functional.py:1805: UserWarning: nn.functional.sigmoid is deprecated. Use torch.sigmoid instead.\n",
            "  warnings.warn(\"nn.functional.sigmoid is deprecated. Use torch.sigmoid instead.\")\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "tensor(0.0060, device='cuda:0', grad_fn=<MseLossBackward>)\n",
            "tensor(0.0054, device='cuda:0', grad_fn=<MseLossBackward>)\n",
            "tensor(0.0050, device='cuda:0', grad_fn=<MseLossBackward>)\n",
            "tensor(0.0045, device='cuda:0', grad_fn=<MseLossBackward>)\n",
            "tensor(0.0046, device='cuda:0', grad_fn=<MseLossBackward>)\n",
            "tensor(0.0049, device='cuda:0', grad_fn=<MseLossBackward>)\n",
            "tensor(0.0045, device='cuda:0', grad_fn=<MseLossBackward>)\n",
            "tensor(0.0044, device='cuda:0', grad_fn=<MseLossBackward>)\n",
            "tensor(0.0047, device='cuda:0', grad_fn=<MseLossBackward>)\n",
            "tensor(0.0051, device='cuda:0', grad_fn=<MseLossBackward>)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C_cx3ErznvHU"
      },
      "source": [
        "#moving images to cpu for plotting\n",
        "noisy_images_cpu = noisy_images.cpu()\n",
        "output_cpu = output.cpu()"
      ],
      "execution_count": 10,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AxGg2_WWnlWq",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 318
        },
        "outputId": "8099d7c1-23be-4575-9d85-dd8ec8b07e11"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "print(losses)\n",
        "plt.plot(losses)\n"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[0.012201098863283794, 0.005636733317871888, 0.0052547811751564345, 0.00505583765146633, 0.00492820870578289, 0.004835634439562758, 0.004747702921306093, 0.004699953449890017, 0.004647523338347673, 0.004606422742456198]\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[<matplotlib.lines.Line2D at 0x7f46a001f090>]"
            ]
          },
          "metadata": {},
          "execution_count": 11
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "G8TP2E1Nnq3g",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 543
        },
        "outputId": "0bf935fd-f334-4684-92f7-c9774e06a27c"
      },
      "source": [
        "noisy_images_cpu = noisy_images_cpu.detach().numpy()\n",
        "noisy_images_cpu = noisy_images_cpu.reshape(-1,28,28)\n",
        "output_cpu = output_cpu.detach().numpy()\n",
        "output_cpu = output_cpu.reshape(-1,28,28)\n",
        "plt.suptitle('Noised Images')\n",
        "for j in range(6):\n",
        "  plt.subplot(2,3,j+1)\n",
        "  plt.xticks([])\n",
        "  plt.yticks([])\n",
        "  plt.imshow(noisy_images_cpu[j],cmap='gray')\n",
        "plt.show() \n",
        "plt.suptitle('Denoised Images') \n",
        "for j in range(6):\n",
        "  plt.subplot(2,3,j+1)\n",
        "  plt.xticks([])\n",
        "  plt.yticks([])\n",
        "  plt.imshow(output_cpu[j],cmap='gray') \n",
        "plt.show() "
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 6 Axes>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWAAAAEHCAYAAACQkJyuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAfKUlEQVR4nO3deZRU9ZUH8O9VtmYXWRQacAFGEUHiBqJiQEVFcY1GHXaNOmbIyCiJcaIGjxF1cI9KjOaoKCoe3DKaIRNExQ0heIAIomiDCCggCAiCy5s/uvh576WrqO509a+W7+ccjvf1rX71ql7Vz/du/xZJkgRERFT3dot9AEREpYoNMBFRJGyAiYgiYQNMRBQJG2AiokjYABMRRcIGmHJORC4Ukek52G8iIl1qe79EdYUNcIkQkQoR2Soim0Rkg4i8ISKXikjOPwNJkjyWJMmJuX4eTURmishFdfmcRNXFBri0nJYkSTMAnQFMAPBLAA/GPSSi0sUGuAQlSfJlkiTPAzgPwHAR6QEAItJQRP5bRJaLyGcicr+IlKVyx4nIChH5TxH5XERWicjIHfsUkRYi8oiIrBGRZSLyXzuurkVkhIjMSsUiIren9rFRRBZk8/yp/FWp510pIqOyfb3q2MepYz9DRE4RkSUi8oWI/Fo9/ggReTN1p7BKRO4RkQYqf6KIvC8iX4rIvSLyir7aFpFRIrJIRNaLyP+KSOddvXYqTWyAS1iSJLMBrABwTOpHEwB0A3AIgC4AOgC4Vv3KXgBapH4+GsDvRWSPVO7uVG4/AP0BDAMwEjs7EcCxqedpAeBcAOt29fwichKAKwGcAKArgOOr+XL3AtBI7fMBAP8K4NDU6/+NiOybeux3AK4A0BpAXwADAfxb6jhaA3gawNUA9gTwPoCjdjyJiJwO4NcAzgLQBsBrAKZk8dqpFCVJwn8l8A9ABYDjq/j5WwCuASAAvgKwv8r1BfBxKj4OwFYA9VT+cwB9AOwOYDuA7ip3CYCZqXgEgFmpeACAJanf2009flfP/xCACSrXDUACoEua1zsTwEXu2HdPbTdL/e6R6vFzAZyRZl//AeCZVDwMwJvuuD9Rz/USgNEqvxuALags+1T52vmvdP/Vq0ZbTcWpA4AvUHm11hjAXBHZkRNUNq47rEuS5Fu1vQVAU1ReKdYHsEzllqX2bSRJMkNE7gHwewCdRWQaKq9sG+3i+dujspHU+6+OdUmSfJeKt6b++5nKb029FohINwC3ATgsdUz11HO3R2WDu+P1JCKyQu2nM4A7RWSi+pkA6JDutSdJsrGar4WKBEsQJUxEDkdlIzkLwFpUNkIHJUnSMvWvRZIkTbPY1VoA36Cy8dmhE4BPq3pwkiR3JUlyKIDuqLySvSqL518FoKPbf67cB2AxgK5JkjRHZUlhx/8VVgEo3/FAqfy/Rbn63U8AXKJeQ8skScqSJHkDSPvaqUSxAS5BItJcRE4F8ASAyUmSLEiS5HtU1kVvF5G2qcd1EJFBu9pf6sryKQA3ikiz1B+dxgKYXMVzHy4iR4pIfVSWHL4G8H0Wz/8UgBEi0l1EGgO47p97FzJqBmAjgM0icgCAy1TufwAcnPojXj0Al6OyvrzD/QCuFpGDUq+hhYj8JBVX+dpz+Dooz7EBLi0viMgmVF6lXYPK22z9h7JfAvgQwFsishHA/wH4lyz3/e+obFQ+QuUV9eOorNt6zVHZ0K5HZRlhHYBbd/X8SZK8BOAOADNSj5mR5XHVxJUALgCwKXWsT+5IJEmyFsBPANySOvbuAOYA2JbKPwPgZgBPpF7DQgAnp34902unEiRJwgnZiWoq1dVuBYALkyR5OfbxUGHhFTBRNYnIIBFpKSIN8UN9+K3Ih0UFiA0wUfX1BbAUlX84PA2V3de2Zv4Vop2xBEFEFAmvgImIImEDTEQUCRtgIqJI2AATEUXCBpiIKBI2wEREkbABJiKKhA0wEVEkbICJiCJhA0xEFAkbYCKiSNgAExFFwgaYiCgSNsBERJFUa1VkEeHclXkiSRLZ9aOyw/OaP3hei9baJEna+B/yCpiIKPeWVfVDNsBERJGwASYiioQNMBFRJGyAiYgiYQNMRBQJG2AiokjYABMRRcIGmIgoEjbARESRsAEmIoqEDTARUSRsgImIIqnWbGhEVDURO4nZbrv9cG2TJHZSsu+//77Kx+mfU9V23313s33qqaeG+KKLLjK5q666KsSLFy/O7YHVEK+AiYgiYQNMRBQJG2AiokhKogas62w6BoBGjRqZbV2H87W7b775JsTfffddjY7F75NqTtddfW3Qn+dvv/02xP4cZDoner/Nmzc3uWOOOSbEJ5xwgsl98MEHIX7iiSdMbs2aNVk9N+2sXj3bZA0fPjzEAwcONLk+ffqEeMmSJSaXL/V2XgETEUXCBpiIKJKiLEE0bNjQbA8YMCDE1113ncl16NDBbG/fvj3EGzZsMLn33nsvxK+//rrJ6dvKsrIyk9OlC30rDACzZ88O8SeffIJ09O12Kd226hJA27ZtTe7CCy8Mcf/+/U3uo48+MtuPPPJIiJcts8tz6c/Lvvvua3IXX3xxiHWXJwBo2rRpiDdu3Ghyb731Voj/9Kc/IZ1SPa81Vb9+fbPdo0ePEDdo0MDkmjVrFmJfosp1CcJ3S0x3bnkFTEQUCRtgIqJI2AATEUUi1ak7iUjeFql0zaW8vNzkXnjhhRDrmhGwc3clX7vRdN1I13UBW+Px9SVdf/K/p4dLTpo0Ke0+vSRJ0h9oNcU+r/o913U7ALj66qtDrOux/rG+e5KvtX/55ZchXr16tcntscceIW7VqpXJ+W6Kmv4bwQMPPGByd999d4hXrlxpcqVyXnPhsMMOM9uzZs0Kse8aes4554R4+vTpJpepG2lt1OWrqAHPTZLkMP84XgETEUXCBpiIKJKi6Yamb/OPPfZYk+vcuXOIfckh02g3X0rQtxWZygz+9kN3ndm6davJLViwIO2xFCv//rRr1y7Ef/jDH0xu0KBBIfbdjHSZwd9S+vO85557hrhly5Zp9/P111+bXEVFRYinTZtmco899liIP/zww7T7pJrzJaBbbrnFbOsuhP4cvPvuuyH23y39GfRd21q3bh1iXboCgC1btqTdp5btd5lXwEREkbABJiKKhA0wEVEkRVMD1rUgXTcEgCZNmoTY12bWrVtntqdOnRpiPfQYsPUfX+M75JBDQnzEEUeYnB5C+8c//tHk9FDkUuFrbqeddlqI/YxWurY+Z84ck3v88cdDrIeQA8CPf/xjs927d+8Q+25oU6ZMCbGexQwA5s2bF2L/WanpjHiUma7PdunSxeT0DGeA/T4/9NBDJqfPV6a/5/i/CfzoRz8K8cKFC01u+fLlVT53TfEKmIgoEjbARESRFGwJws9utPfee4e4Z8+eJqe7JPnSgZ8s+6abbgrx2rVrTU7fxviuVM8880yIdTkCsCWQmTNnmpwfGVcs/PujR6rttddeJjd27NgQ+25HulwwdOhQk1u6dGna53/44YfNtj4HmzdvNrlt27aFONMCmqXSTbCu+c+KLidedtllJudnGtTdxHw3Qf3d8s+htzdt2mRyeia7r776yuSy/QxwNjQiojzHBpiIKBI2wEREkRRsDdjPfqVnPmrfvr3J6eG/vpvRgw8+aLb1yha+m5Gu4/gatB7COn/+fJPTXaR0vbGY+ffnwAMPDLFflWSfffYJsX9/dE3e13wzLbTpa/2+lqfp3822dke1x7/neti47qJYlblz54Z4xYoVWT+nru37z5zerunKGRyKTESU59gAExFFUlAlCN2drGvXriY3YsSIEOsJtgFbHvDdWHr16mW29YxK1SkX6FuVmnZdKXS67OC7AuoJyw866KC0+/Ajj/785z+H2N8OVud9zfaxdX2uuCjnzuXE448/PsS+y6IvLenZ8/xMg9m+n7leoDMTXgETEUXCBpiIKBI2wEREkRRUDVjPonXUUUeZXJs2bULsV0No2rRpiLt162Zyt99+u9nu3r17iF9//XWT0zMh6RiwCzTGrCnVJd99SM84NnnyZJPbb7/9QuzfHz3c+M477zS5Tz/9NO3vZVIo9dRCOc5catGihdkeN25ciP3Mef5vBH/9619DXIjfO14BExFFwgaYiCiSgipB6NmN9OxjgJ30/LzzzjM53WXNL+yoyxMAcOmll4b47LPPNjl9Wz1x4kSTK8VbSf/e3X///SH23QQ1311Id18bNWpU2tzTTz+ddj+FePtZynSZUHc7A2yZ0H+vbrvtNrPtF80sNLwCJiKKhA0wEVEkbICJiCKR6tQuRSRvC516OGOnTp1M7vTTTw+xXnAPADp37my2Dz744BA3btzY5PSw2PPPP9/k/KKQuZYkiez6Udmp6XnVq0wAwIwZM0LsVwXRtVw/nNQdS9ptP5Pd9ddfH2LdHQnYeTh4ppnT8kk+nNe6oFe9eOWVV0xOL2rrZzjz31+/ak0em5skyWH+h7wCJiKKhA0wEVEkbICJiCIpqH7AmegaX0VFhcndc889IW7evLnJXXDBBWb70EMPDbGfJu/zzz8PMfud7tyfV69eoftTA3ZKUN9387XXXguxn8ZSr3SihzMDwJQpU0K8ceNGk3v22WfNtp62cN68eSan+5fzvOaGr+3rfuL+7wWaX7Fm/fr1tXtgkfEKmIgoEjbARESRFE0JQvO3kXrb36pu3rzZbOvuMZ6eAS2fuzLVFf8e/OUvfwnxq6++anK6BOHPjy4B+FvV8ePHh9gv5jls2LAQ+xm1Ro4cabZ/+tOfhth3WbvkkktCrMtMVR0r1YxfpHXMmDEh9t85/Z30Uw74hXILHa+AiYgiYQNMRBQJG2AiokiiDEXWdT5f89PHU1t1Vv0cfvjso48+arb1sGU/nHXgwIEhnjNnjsnVda0wH4es+nOp1fRc6n36WqHuvjR06FCT80PF9UrZ/lz9/e9/D7E+/wCwatWqENdF3T8fz+s/8fwhbteuncktWLAgxK1btza5TEP+C3jFcQ5FJiLKJ2yAiYgiidINTa9K4bsP6VsMf7tRU7oLTN++fU3Oz8avb2n07F4AsGjRohCze9LOaqPM4Olz52ecmz17doh1GQEA7r33XrN93333hbhPnz4mpxcT9V3dLr/88hBnmsWNdqZXvRg8eLDJ6e/9tm3bTE6vPONHWxZQySErvAImIoqEDTARUSRsgImIIqmTbmi+xtevX78Q6yGJAPDOO++EWM92BdjZ7/2QRF1vKisrMzk9DPXmm282uWbNmpntDRs2hPjEE080OV1njF0DLqbuSvrc+RnPdNezTz75xOT08Gb/OfafuTZt2oR46tSpJnf44YeHeN26dSY3ZMiQEOvPZlXPWRuK6bzqFWVeeuklkzv66KND/N5775ncgAEDQrxmzZocHV2dYzc0IqJ8wgaYiCiSOumG5m8HTz755CpjADjppJNC/POf/9zkFi9eHGI9QgmwI530LSVgbz/17S5gSw4AMHbs2BC/++67Jhe77FAK/Pk44IADQjx8+HCTW7p0aYiXLFlicr5coUsZfqJ9rVWrVmZbz6qWaSJ32lm3bt1C7Cfa113Ppk2bZnKbNm3K7YHlEV4BExFFwgaYiCgSNsBERJHUSQ3Y1131jPe+rqpnK/NdxDp27Jj2OXSd2XcP0kNY9XBiALjyyivN9ssvvxxiDj2tG/p8ffHFFyb39ttvh1h3QwSAiRMnhlh3awKALVu2mG09pLW8vNzk9HBn371Rd6XKNGTay/R5LFb+/dHD/n3X0NWrV4fYz0joh5wXM14BExFFwgaYiCiSOilB+Ns6PcLNd/vp0aNHiP0kzvXr1w+xn5x7xYoVIfZd1PQonOnTp5ucv61lV7O6l2kSfn0+fFezn/3sZyG+9tprTU6PpgJ2Ljtoeta9iooKk9Of1ZqWpDItOlDMGjVqFGJfhtQlIT/jWSl9B3kFTEQUCRtgIqJI2AATEUUSZVFOzdeG/LaW6VgzLfSpa3fFUn8rplmzaoM/53rVFcDWgPUKGID9W4NfbHX+/Pkh9rXKbI+nOp+5Qj6v/hwcfPDBIX7uuedMTv+dxi+8uWzZshwcXd2qou7P2dCIiPIJG2AiokiilyCoZgr5VpXSK9bzqkcbAnZGOj+rXJF2Q2MJgogon7ABJiKKhA0wEVEkdTIUmYhKm5+OwG+XKl4BExFFwgaYiCgSliCIiLKU7QjHbGfA4xUwEVEkbICJiCJhA0xEFEl1a8BrART+VEWFr3Mt74/nNT/wvOa5bKduqOJxVZ7bas0FQUREtYclCCKiSNgAExFFwgaYiCgSNsBERJGwASYiioQNMBFRJGyAiYgiYQNMRBQJG2AiokjYABMRRcIGmIgoEjbARESRsAEmIoqEDTARUSRsgImIImEDTEQUCRtgIqJI2AATEUVSrTXhRITrF+WJJEmktvbF85o/eF6L1tokSdr4H/IKmIgo96pcHJUNMBFRJGyAiYgiYQNMRBQJG2AiokjYABMRRcIGmIgoEjbARESRsAEmIoqkWiPhiIhiErEDBXfb7YdryFatWqX9vQ0bNpjtb775pnYPrIZ4BUxEFAkbYCKiSFiCoJLQuHHjEPfs2dPkFi5cGOLNmzfX2TEVO10u2H333U2uadOmIe7du7fJnXvuuSHu06ePyXXq1Mlsl5WVpX0Obfny5WZ7+PDhIX7zzTdNLknqbg4jXgETEUXCBpiIKBI2wEREkUh16h2xJ3iuV++HknWDBg1MbuvWrSGu5mtKu+33U5e1oV3hxN2Z1a9f32zfcMMNIb788stNbtq0aSEeM2aMyX355Zc5OLr0Cvm8tmvXzmyfeeaZIR48eLDJ9erVK8Rt27Y1OX3udDezqnz//fdVxoD9Lvv9zJ49O8QnnXSSyfkua7VkbpIkh/kf8gqYiCgSNsBERJHkdTc0XXIA7K3j6NGjTW7s2LEhnjlzpsl99913Ifa3piNHjjTbRx55ZIhfffVVk5s6dWqIv/rqq0yHTpH58zxgwIAQ6y5pANCvX78Q77vvvia3YMGCEOvPEVXS3ckmT55scscee2yI/fnQfOng66+/Tptbv3692f7www9DvGyZXfWnf//+IS4vLze59u3bh9h3g3vttddC/O2336Y97trAK2AiokjYABMRRcIGmIgokryuATdp0sRs6xrwfvvtZ3Ljx48P8XnnnWdya9euDXHnzp1NTteOAVsDPO6440xOD1lcsmSJyeVTFzUC9tprL7PdpUuXEPuZsBYtWhTidevW5fbAisz+++8f4kMOOcTk9NBg3U0UAObPnx/ip556yuT0sOE1a9aYXEVFhdnWf4vx5/zAAw9Mm9u0aVOIGzVqZHK6y5rvplrb33NeARMRRcIGmIgokrwuQeyxxx5mu2XLliH2I1t0d5g999zT5PTtxgEHHGByfhJn3V3Gl0AaNmyYzWFTBL7Lou9e2KJFixBv2bLF5CZNmhTizz77zOR8NyiydElAjygE7Pfw4YcfNrlZs2aFeOPGjSan3/NME7AD9jupR94BwEEHHRRiX3bSXebeeOMNk9Ndz3JdWuQVMBFRJGyAiYgiYQNMRBRJ3tWAdc2na9euJqfrvJ6ewWj16tUmp+s/fhY13wVF13x07djvl93O8ov/bFx88cVmW9cOdRcowA493b59ew6Ornjp78i4ceNMTn/v9PBioOa1dT+k+ayzzgrxFVdcYXK6Pjx37lyTmzJlSoj997wu6/68AiYiioQNMBFRJHlXgtAlAn17AdiRNf42Yd68eSH2M5Xpx+qRO8DO3Zc0PYIO4O1pvtFlhRNOOMHkWrdubba3bdsW4gkTJpicXogz1yOfvEwLABQC/d2qrcnr9Xuiu54CwJAhQ8y2HgHbrFkzk9PHc8stt5jcypUrQxyzqyGvgImIImEDTEQUCRtgIqJI8q4GrIcv9uzZ0+R0DdivTqBnytePA2x3lO7du2d9LB06dDDbZWVlIc7Rwn2Uga/P6qHiv/rVr0zOD1l9//33Q+xXTIlZAyzEum9t8+dVz1x20003mdwpp5xitvUQc13nB+xqJkuXLs34nLHwCpiIKBI2wEREkUQvQfhbAT1xs19kTy+Y6XXq1CnEehYkwM6A5mdD84vu6ZE2fnSVnszdj7Zjd6Xc86WlI444IsS+tOTfo7vuuivEvptiqb6fdcmfu7333jvEV111lckNGzYsxM2bN8+4X10+8qUkvfCmXiAUAFasWBFiP1Oa/jxwNjQioiLFBpiIKBI2wEREkUSvAfsaix4W+s4775jcOeecE2I/K9Kpp55aZQzYIYr++Xx3Jb3thzYeffTRIZ4zZ47J+VpybWOdEmjcuLHZ1l3P/Kx2X3zxhdl+8cUXQ8xVLuqG/rtF3759Te7xxx8Pse/uqb+D/nPvZ1XTXc/8TIf670K/+c1vTE6vwvHkk0+m3Weu8QqYiCgSNsBERJFEL0F4+lZeT5QN2ImT/YKdenFN3+Ul3f6Bnbsk6dnR/H50d5mFCxea3MsvvxziuryFKXb6fBx11FEm57sWaX/729/M9po1a2r3wGgnvkup7mr2wAMPmJwuO/gygy4fzZ492+Sef/55s63LFX369DE5PWpOj5gDgNGjR4dYf3cB20WtpqW/bGfV4xUwEVEkbICJiCJhA0xEFEne1YB1reSjjz4yOV0THjRokMnpmosezgwAH3/8cYjffvttk/NDikeMGBFiXcMC7Oz811xzjcnp2dlmzJiRNkfVo4eDX3fddSanuyL6urseegzwHMSg67N6xRrALozqF8ycPHlyiD///HOT8+dR/53mscceMzn9N6OLLrrI5PR0BR07djQ5XQOuqWxrx7wCJiKKhA0wEVEkbICJiCLJuxqw5ldZ1UMG+/fvb3K6FvTQQw+ZnJ5V3+9TTzEJAEOHDg2xr+Pobd+fWG/74c2sP2bPv3eHH354iHv16mVyuu7vVzz4xz/+kYOjq32FMM2oPifVWTX6s88+C/GoUaPSPs5PB1md74v+3ulpDAD7NyTfp1+vbpPraQQy4RUwEVEkbICJiCLJ6xKEvxXRs6N98MEHJqdLCzfeeKPJ+ZmxNL/qhR7i7G9bdLeWRx55xOR0Vxp/S0XZ87PcjRw5MsR+xjP9+Zg0aZLJ6dmugPy9vc+X49JlBj/rXI8ePUKsF80FbJlh0aJFJqdnLvO3+bl43f7zcfbZZ4fYl062bNkS4uXLl+f82NLhFTARUSRsgImIImEDTEQUSV7XgD09pPj00083OV1jWrduncnpFRB8Lch3S9PDmP2KrLre5YdW+i4wlD19TsrLy03u+OOPT/t7+nw888wzJsdVL6pHf9bHjRtnchdccEGI/SoxeprPO+64w+T0qhf++1FbdVY9Xemll15qcoceemja33viiSdC7Ic71yVeARMRRcIGmIgokoIqQehuR6tWrTK5bEcU+ZwvQegVMvxj27RpE2J967Or56TMdHe/M8880+T0Sga+W6JeHcHfRvJ8VI9+b/2te/v27UPsu2bqbpx+4cuuXbuGWJcjAGDx4sUh1l3CgMznznc1O//880N87bXXmpz+jn766acmd+utt4Y4ZrmKV8BERJGwASYiioQNMBFRJAVVA86kpjW/DRs2mO1HH300xL6mpVfE0LN0AbZbGmc/q54mTZqE+MILLzQ5XXP0Q0Zvv/32EG/fvj1HR1cadDexp59+2uT0atT6XAF26Hjr1q1NTg8j1ysUA/bcPfvssyanhzC3bdvW5CZMmGC2Bw8eHGJfH9a15SuuuMLkVq5ciXzAK2AiokjYABMRRSLVuXUXkaLr2+NHxu2///4h9otrdujQIcTTp083uXPPPTfEeta0XEmSRHb9qOzU9Xn1k64PGzYsxHfffbfJ6RLE9ddfb3J64U1921rI8uG8+jKDXtDyF7/4hcnp74Q/r/q75UtEa9euDXFFRYXJ6a6H++yzj8n52Qs1PwJ2zJgxIfZllQiTsM9NkuQw/0NeARMRRcIGmIgoEjbARESRlFwNeFeLCjZs2DDEvh6pu9X44ZODBg0K8ezZs00uF0Md86FWWFOtWrUy22+++WaIu3TpYnJ6cU3d5Qiww0uLZfazfDyvDRo0CHG/fv1MTq9E4uu1un7vv3c15WdV0wv13nDDDSanu5rFXHgzhTVgIqJ8wgaYiCiSohkJl61dlVx0dxm/8OaQIUNCrEfFAcBtt90WYj2BNQAsW7Ys6+cvVvoWtHfv3ibXsWPHEPtSwsyZM0Osuy5V9VjKDf2d0OcDAAYOHBji3/3udyZ3xhlnhFiX9gB77vwituvXrw/xiy++aHL+OXSZwY9ALYTvGq+AiYgiYQNMRBQJG2AiokhKohtatqtleH7Y48SJE0M8atQok9M1LT3TEwCMHz8+xHrRz+oej/u9vOuulC2/2Ono0aND7Gepeu6550K8bds2kyuEGt+u+M9mIZ9X39XMD03WMp27Iq3tsxsaEVE+YQNMRBRJSXRDq+mtqh9189vf/jbEfuFCPYtaWVmZyelJq/2sXcVwG11dGzduNNu+ZEOFyX+WuTDBrvEKmIgoEjbARESRsAEmIoqkJLqh5UK9erZ8Xl5eHuJVq1aZnB7KWVs130LurkTp8bwWLXZDIyLKJ2yAiYgiKYluaLngJ3jO9YxnNR3NR3Fke770aLEiHQFGGfAKmIgoEjbARESRsAEmIoqkujXgtQCW7fJRJSjXdVm3/861vHue11qW7efB1X15XotXlee2Wv2AiYio9rAEQUQUCRtgIqJI2AATEUXCBpiIKBI2wEREkbABJiKKhA0wEVEkbICJiCJhA0xEFMn/A1wtc7U4/QQiAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 6 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pE-s2xFzjZ2D",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "23eecf21-1579-4cc2-ddd7-1a9a29ac9638"
      },
      "source": [
        "#finding results on testset\n",
        "with torch.no_grad():\n",
        "  for img,_ in testset:\n",
        "    noisy_img = add_noise(img)\n",
        "    noisy_img = noisy_img.to(device)\n",
        "    reconstructed = model(noisy_img)"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/torch/nn/functional.py:1805: UserWarning: nn.functional.sigmoid is deprecated. Use torch.sigmoid instead.\n",
            "  warnings.warn(\"nn.functional.sigmoid is deprecated. Use torch.sigmoid instead.\")\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eyp9NpdyoGzD"
      },
      "source": [
        "img = img.cpu()\n",
        "noisy_img = noisy_img.cpu()\n",
        "reconstructed = reconstructed.cpu()"
      ],
      "execution_count": 14,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Nl4f4wt9jyzY",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 806
        },
        "outputId": "6a75eb33-adf9-4dc2-93fc-c78b9126230e"
      },
      "source": [
        "img = img.detach().numpy()\n",
        "img = img.reshape(-1,28,28)\n",
        "noisy_img = noisy_img.detach().numpy()\n",
        "noisy_img = noisy_img.reshape(-1,28,28)\n",
        "reconstructed = reconstructed.detach().numpy()\n",
        "reconstructed = reconstructed.reshape(-1,28,28)\n",
        "plt.suptitle('Original Images')\n",
        "for p in range(6):\n",
        "  plt.subplot(2,3,p+1)\n",
        "  plt.xticks([])\n",
        "  plt.yticks([])\n",
        "  plt.imshow(img[p])\n",
        "plt.show() \n",
        "plt.suptitle('Noised Images') \n",
        "for p in range(6):\n",
        "  plt.subplot(2,3,p+1)\n",
        "  plt.xticks([])\n",
        "  plt.yticks([])\n",
        "  plt.imshow(noisy_img[p])\n",
        "plt.show()\n",
        "plt.suptitle('Denoised Images')\n",
        "for p in range(6):\n",
        "  plt.subplot(2,3,p+1)\n",
        "  plt.xticks([])\n",
        "  plt.yticks([])\n",
        "  plt.imshow(reconstructed[p])  \n",
        "plt.show()  "
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 6 Axes>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 6 Axes>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 6 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    }
  ]
}