AI engineering glossary · Computer vision
What is bottleneck block?
Bottleneck block: A residual block that shrinks channels with a 1 x 1 convolution, does a 3 x 3 convolution at the smaller width, then expands with another 1 x 1 convolution. It gives deep networks like ResNet-50 lower cost per layer.
Bottleneck block: an example
A ResNet-50 block might go 256 to 64 channels, apply a 3 x 3 convolution, then go back to 256.
Related terms
- 1x1 convolution
- ResNet
- Residual connection
More Computer vision terms
- Bounding box
- Canny edge detector
- Channel
- CIFAR-10
- CLS token
- Colour jitter
- Confidence score
- Connectionist temporal classification
Where you learn it
Bottleneck block is part of the AI Engineering Program, which teaches it with notes, a hands-on lab and an exam. The first lesson is free.