0PricingLogin▼
Deep Learning Academy · Lesson

Stride, Padding & Pooling

Control output size and downsample.

Output Shrinks by Default

When a 3x3 kernel slides over an image, the edges have no room. So the output feature map comes out a little smaller than the input.

Padding Adds a Border

Padding wraps the image in a ring of zeros so the kernel can reach the corners. Now the output can stay the same size as the input.

All lessons in this course

  1. Convolution: Kernels Slide Over Pixels
  2. Stride, Padding & Pooling
  3. Channels, Feature Maps & Receptive Fields
  4. Assemble a CNN Image Classifier
← Back to Deep Learning Academy