AI Blog

AI Blog

by Michele Laurelli

Feature Map

/ˈfiːtʃər mæp/
Concept
Definition

The output of applying a convolution filter to an input, representing detected features.

Each filter produces one feature map showing where specific features (edges, textures, patterns) are detected. Multiple filters produce multiple feature maps, each capturing different aspects of the input.

Examples

1

Edge detection feature map

2

Texture feature maps

3

High-level semantic features

Michele Laurelli - AI Research & Engineering