Concept Lab · Week 14 · Other Modalities
Convolution kernels
Slide a 3×3 kernel over an image and see edges appear.
The idea: Convolution and filters
A convolution slides a small grid of weights (a kernel, e.g. 3×3) over the image. At each position it multiplies the kernel with the patch underneath and sums the result into one output pixel.
Hand-designed kernels do classic image processing: a blur averages neighbours; a Sobel kernel responds to horizontal or vertical edges; sharpening amplifies the centre relative to neighbours. Try them in the simulation.
CNNs (week 17) learn their kernels from data instead: early layers learn edge detectors that look much like Sobel; deeper layers learn textures, parts and objects.
Next simulation: Backpropagation, node by node