AI engineering glossary · Computer vision
What is 1x1 convolution?
1x1 convolution (also called Pointwise convolution): A convolution whose kernel covers a single pixel but all input channels. It mixes information across channels at each location, and is used to cut or raise the channel count cheaply, as in bottleneck blocks and Inception modules.
1x1 convolution: an example
A 1 x 1 convolution can compress a 56 x 56 x 256 feature map to 56 x 56 x 64 before an expensive 3 x 3 convolution.
Related terms
More Computer vision terms
- Affine transformation of images
- AlexNet
- Anchor box
- Aspect ratio
- Average pooling
- Binary mask
- Bottleneck block
- Bounding box
Where you learn it
1x1 convolution is part of the AI Engineering Program, which teaches it with notes, a hands-on lab and an exam. The first lesson is free.