Fix segmentation example (#876)

* removed torchvision model and added custom model

* minor fix

* Fixed relative imports issue

Author: akshaykulkarni07

pl_examples/full_examples/semantic_segmentation/models/unet/__init__.py

@@ -0,0 +1,4 @@
+# For relative imports to work in Python 3.6
+import os
+import sys
+sys.path.append(os.path.dirname(os.path.realpath(__file__)))

pl_examples/full_examples/semantic_segmentation/models/unet/model.py

@@ -0,0 +1,42 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+
+from parts import DoubleConv, Down, Up
+
+
+class UNet(nn.Module):
+    '''
+    Architecture based on U-Net: Convolutional Networks for Biomedical Image Segmentation
+    Link - https://arxiv.org/abs/1505.04597
+    '''
+    def __init__(self, num_classes=19, bilinear=False):
+        super().__init__()
+        self.bilinear = bilinear
+        self.num_classes = num_classes
+        self.layer1 = DoubleConv(3, 64)
+        self.layer2 = Down(64, 128)
+        self.layer3 = Down(128, 256)
+        self.layer4 = Down(256, 512)
+        self.layer5 = Down(512, 1024)
+
+        self.layer6 = Up(1024, 512, bilinear=self.bilinear)
+        self.layer7 = Up(512, 256, bilinear=self.bilinear)
+        self.layer8 = Up(256, 128, bilinear=self.bilinear)
+        self.layer9 = Up(128, 64, bilinear=self.bilinear)
+
+        self.layer10 = nn.Conv2d(64, self.num_classes, kernel_size=1)
+
+    def forward(self, x):
+        x1 = self.layer1(x)
+        x2 = self.layer2(x1)
+        x3 = self.layer3(x2)
+        x4 = self.layer4(x3)
+        x5 = self.layer5(x4)
+
+        x6 = self.layer6(x5, x4)
+        x6 = self.layer7(x6, x3)
+        x6 = self.layer8(x6, x2)
+        x6 = self.layer9(x6, x1)
+
+        return self.layer10(x6)

pl_examples/full_examples/semantic_segmentation/models/unet/parts.py

@@ -0,0 +1,68 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+
+
+class DoubleConv(nn.Module):
+    '''
+    Double Convolution and BN and ReLU
+    (3x3 conv -> BN -> ReLU) ** 2
+    '''
+    def __init__(self, in_ch, out_ch):
+        super().__init__()
+        self.net = nn.Sequential(
+            nn.Conv2d(in_ch, out_ch, kernel_size=3, padding=1),
+            nn.BatchNorm2d(out_ch),
+            nn.ReLU(inplace=True),
+            nn.Conv2d(out_ch, out_ch, kernel_size=3, padding=1),
+            nn.BatchNorm2d(out_ch),
+            nn.ReLU(inplace=True)
+        )
+
+    def forward(self, x):
+        return self.net(x)
+
+
+class Down(nn.Module):
+    '''
+    Combination of MaxPool2d and DoubleConv in series
+    '''
+    def __init__(self, in_ch, out_ch):
+        super().__init__()
+        self.net = nn.Sequential(
+            nn.MaxPool2d(kernel_size=2, stride=2),
+            DoubleConv(in_ch, out_ch)
+        )
+
+    def forward(self, x):
+        return self.net(x)
+
+
+class Up(nn.Module):
+    '''
+    Upsampling (by either bilinear interpolation or transpose convolutions)
+    followed by concatenation of feature map from contracting path,
+    followed by double 3x3 convolution.
+    '''
+    def __init__(self, in_ch, out_ch, bilinear=False):
+        super().__init__()
+        self.upsample = None
+        if bilinear:
+            self.upsample = nn.Upsample(scale_factor=2, mode='bilinear', align_corners=True)
+        else:
+            self.upsample = nn.ConvTranspose2d(in_ch, in_ch // 2, kernel_size=2, stride=2)
+
+        self.conv = DoubleConv(in_ch, out_ch)
+
+    def forward(self, x1, x2):
+        x1 = self.upsample(x1)
+
+        # Pad x1 to the size of x2
+        diff_h = x2.shape[2] - x1.shape[2]
+        diff_w = x2.shape[3] - x1.shape[3]
+
+        x1 = F.pad(x1, [diff_w // 2, diff_w - diff_w // 2, diff_h // 2, diff_h - diff_h // 2])
+
+        # Concatenate along the channels axis
+        x = torch.cat([x2, x1], dim=1)
+        return self.conv(x)

pl_examples/full_examples/semantic_segmentation/semseg.py

@@ -10,9 +10,9 @@
 import torchvision
 import torchvision.transforms as transforms
 from torch.utils.data import DataLoader, Dataset
-from torchvision.models.segmentation import fcn_resnet50
 
 import pytorch_lightning as pl
+from models.unet.model import UNet
 
 
 class KITTI(Dataset):
@@ -128,9 +128,7 @@ def __init__(self, hparams):
         self.root_path = hparams.root
         self.batch_size = hparams.batch_size
         self.learning_rate = hparams.lr
-        self.net = torchvision.models.segmentation.fcn_resnet50(pretrained=False,
-                                                                progress=True,
-                                                                num_classes=19)
+        self.net = UNet(num_classes=19)
         self.transform = transforms.Compose([
             transforms.ToTensor(),
             transforms.Normalize(mean=[0.35675976, 0.37380189, 0.3764753],
@@ -147,7 +145,7 @@ def training_step(self, batch, batch_nb):
         img = img.float()
         mask = mask.long()
         out = self.forward(img)
-        loss_val = F.cross_entropy(out['out'], mask, ignore_index=250)
+        loss_val = F.cross_entropy(out, mask, ignore_index=250)
         return {'loss': loss_val}
 
     def configure_optimizers(self):