Skip to content
ML to LLMs
Labs / Neural Network Lab

Convolution and pooling

Step a 3×3 kernel over an image, see each multiply-and-add, then max-pool the feature map.

Theory and worked examples: Convolutional networks

Convolution and pooling

Click any output cell (or use the slider) to see the 3×3 window that produced it.

input image[7, 7]
kernel[3, 3]
feature map[5, 5]
2×2 max pool[2, 2]

Kernel, feature map and pool: teal is positive, orange is negative, stronger colour means a larger magnitude. The purple outline is the current window and the output cell it produces.

Y[1,2] = -1·0 + 0·0 + 1·9 + -1·0 + 0·0 + 1·9 + -1·0 + 0·0 + 1·9 = 27

The same 9 weights slide over every position (weight sharing). Large positive values mean the patch matches the kernel's pattern; that is what a learned filter detects.