Convolutional Neural Network
Also called CNN, ConvNet
A network that scans small patches across an input to detect local patterns, great for images.
Think of it like
Sliding a magnifying glass across a photo, noting edges and textures wherever it lands.
Example
A CNN powers a phone’s photo classifier, spotting edges, then shapes, then whole objects layer by layer.
How it actually works
CNNs apply small learnable filters that slide across the input, reusing the same weights everywhere so they detect a pattern wherever it appears. Stacked layers build from simple features (edges) to complex ones (faces). They dominated computer vision for years and remain strong there, though transformers have encroached on vision too. For text and long-range reasoning, transformers took over.
For product teams
The workhorse behind most image-recognition features of the last decade.
For engineers
A network of shared-weight sliding filters exploiting spatial locality and translation invariance.
Related
- Deep learning — A pillar of deep learning.
- Recurrent Neural Network — For sequences, compare the recurrent neural network.
- Transformer — Largely supplanted for text by the transformer.
Read anything AI without the jargon
Look up any term in plain English, or save terms as you read with the free Chrome extension.
Open DecoderAdd to Chrome