Strided convolution
A strided convolution is a convolution whose filter moves across input positions by more than one step, typically reducing spatial or temporal resolution.
Deep Learning with R, Third Edition
A strided convolution is a convolution that moves its kernel by more than one input location at a time to reduce spatial resolution.
- How it is used
- Uses convolutional downsampling when segmentation or detection depends on retaining spatial location.
- How deep it goes
- Explains the practical role and architecture placement, but does not derive convolution-stride mechanics mathematically.
- Examples
- Demonstrates strided Conv2D in an encoder-decoder segmentation model and a YOLO detection head.
- What you can do after
- Apply it via Keras Conv2D layers to downsample feature maps in location-sensitive vision models.
Read how Deep Learning with R, Third Edition explains strided convolution →
Inside Deep Learning: Math, Algorithms, Models
A convolution performed with stride greater than one so it reduces spatial resolution instead of using a separate pooling layer.
- How it is used
- Replaces max pooling when shrinking images or feature maps, including in a fully convolutional U-Net.
- How deep it goes
- Describes its downsampling role and proposes comparisons, but provides no implementation details or performance results.
- Examples
- Compares a U-Net with pooling replaced and CIFAR-10 networks using standard or anti-aliased strided convolution.
- What you can do after
- Apply it by replacing pooling layers with strided convolutions in a U-Net and comparing performance.
Read how Inside Deep Learning: Math, Algorithms, Models explains strided convolution →
What to know first
- bounding boxA bounding box is a coordinate-defined rectangular region in two or more dimensions that encloses or localizes an object, area, or set of points.
- convolutional layerA convolutional layer is a neural-network layer that applies shared learned filters to local regions of structured input to produce feature representations.
- max poolingMax pooling is an aggregation operation that represents each group of values by its maximum, often reducing the size of a feature representation.
- linear layerA linear layer is a neural-network layer that applies a learned affine transformation to its input, typically using weights and a bias.
Related concepts
- max poolingcontrasts with · Max pooling is an aggregation operation that represents each group of values by its maximum, often reducing the size of a feature representation.
- downsamplingimplements · Downsampling is the reduction of data quantity, spatial resolution, temporal resolution, or sampling rate by retaining, selecting, or aggregating fewer representative values or observations.
- strideuses · Stride is the step or spacing between successive positions in an array access or successive placements of an operation over input data.
Questions about strided convolution
What is strided convolution?
A strided convolution is a convolution whose filter moves across input positions by more than one step, typically reducing spatial or temporal resolution.
Which Manning books explain strided convolution?
Deep Learning with R, Third Edition and Inside Deep Learning: Math, Algorithms, Models.
What should I know before learning strided convolution?
Manning books that teach strided convolution usually introduce bounding box, convolutional layer, max pooling and linear layer first.
What is the difference between strided convolution and max pooling?
A strided convolution is a convolution whose filter moves across input positions by more than one step, typically reducing spatial or temporal resolution. Max pooling is an aggregation operation that represents each group of values by its maximum, often reducing the size of a feature representation.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.

