Convolutional Neural Networks

CNN (Convolutional Neural Networks) are a class of deep learning models that are particularly effective for processing data with a grid-like topology, like images. They are preferred over fully connected networks for image data due to their ability to capture spatial hierarchies and reduce the number of parameters.

CNN filters share weights across different locations in the input assuming input is stationary. We utilize learned kernels in convolutional layers to extract features from the input data.