AI Glossary

Convolution kernel

Also called convolution filter

What does “Convolution kernel” mean?

Convolution kernel: A small set of weights applied across an input to detect local patterns.

Convolution kernel visual guide: A small set of weights applied across an input to detect local patterns.
Visual guide Open full size (new tab)

What is an example of “Convolution kernel”?

A vision model learns filters that respond to edges in image patches.

Keep learning

See also

Deep learningTermTrainingTermOverfittingTermReceptive fieldTermPooling layerTerm

Same topic

Curriculum learningTermContrastive learningTermData augmentationTermContinual learningTermData labelingTerm