A neural network that uses only convolution layers to predict every pixel.
FCN treats a photo like a coloring book page. It colors every tiny square: cat here, couch there.
You meet it in medical scans and self-driving car vision. It gives pixel positions, not one big label.
CNN
FCN is a dense prediction version of CNN. It outputs a pixel map.
Image Segmentation
FCN helped image segmentation move from whole-image labels to pixel labels.
U-Net
U-Net follows the fully convolutional idea and improves detail blending.