CNNs · Level 2 of 5
Convolution
An operation combining a signal with a kernel over shifted positions.
Many neural libraries call cross-correlation convolution even though the kernel is not flipped.
Example
A small learned filter computes a response at each image location.
Listen to the definition and example
Audio transcript
Convolution. An operation combining a signal with a kernel over shifted positions. Many neural libraries call cross-correlation convolution even though the kernel is not flipped. For example: A small learned filter computes a response at each image location.
Explore this concept
Why it matters
This helps you interpret feature shapes and spatial behavior in image networks.
Start with
Quick recall question
Try answering before looking back at the definition.