Scarlett I want to understand what semantic segmentation means in computer vision. How does it classify each pixel in an image into different object categories? Can someone also explain where it is commonly used in real-world applications?