# Convolutional Neural Network Here you want a machine to learn a specific kernel. So instead of handing it a Gaussian, instead you try and have it learn which kernels are going to pull out some meaningful information from an image. ### From FastAI Each convolution filter (kernel) is *learned*. We can think of these as being edge detectors, detectors of higher level objects. So really we can think of convolution filters/kernels as learned **features**. --- Date: 20210902 Links to: [AI MOC](AI%20MOC.md) [Convolution](Convolution.md) Tags: References: * [3b1b idea](https://youtu.be/8rrHTtUzyZA?list=PLwcClAaLqrJyzeqgdc9dJYRh9OYawfkSL&t=1254) * [course22p2/nbs/07_convolutions.ipynb at master · fastai/course22p2 · GitHub](https://github.com/fastai/course22p2/blob/master/nbs/07_convolutions.ipynb)