Semantic Segmentation

Introduction

Semantic segmentation assigns class labels to each pixel in an image, enabling pixel-level understanding.

U-Net

import torch
import torch.nn as nn

class DoubleConv(nn.Module):
    def __init__(self, in_ch, out_ch):
        super().__init__()
        self.conv = nn.Sequential(
            nn.Conv2d(in_ch, out_ch, 3, padding=1),
            nn.BatchNorm2d(out_ch),
            nn.ReLU(inplace=True),
            nn.Conv2d(out_ch, out_ch, 3, padding=1),
            nn.BatchNorm2d(out_ch),
            nn.ReLU(inplace=True)
        )

    def forward(self, x):
        return self.conv(x)

class UNet(nn.Module):
    def __init__(self, in_channels, num_classes):
        super().__init__()
        self.enc1 = DoubleConv(in_channels, 64)
        self.enc2 = DoubleConv(64, 128)
        self.enc3 = DoubleConv(128, 256)
        self.enc4 = DoubleConv(256, 512)

        self.pool = nn.MaxPool2d(2)
        self.up4 = nn.ConvTranspose2d(512, 256, 2, stride=2)
        self.dec4 = DoubleConv(512, 256)
        self.up3 = nn.ConvTranspose2d(256, 128, 2, stride=2)
        self.dec3 = DoubleConv(256, 128)
        self.final = nn.Conv2d(128, num_classes, 1)

Segmentation Models

# Using segmentation_models_pytorch
import segmentation_models_pytorch as smp

model = smp.Unet(
    encoder_name="resnet34",
    encoder_weights="imagenet",
    in_channels=3,
    classes=10
)

Mask R-CNN

import torchvision
from torchvision.models.detection import maskrcnn_resnet50_fpn

model = maskrcnn_resnet50_fpn(pretrained=True)
model.eval()

# Get masks
outputs = model(img_tensor)
masks = outputs[0]['masks']

DeepLab

import torchvision
from torchvision.models.segmentation import deeplabv3_resnet50

model = deeplabv3_resnet50(pretrained=True)
model.eval()

# Get segmentation
output = model(img_tensor)['out']

Practice Problems

1. Build U-Net architecture
2. Train on segmentation dataset
3. Use pre-trained segmentation models
4. Evaluate with IoU
5. Visualize segmentation masks