Explore Library
Code QuizIntermediate

Convolution Output Size Formula

Find the mistake in this convolution output-size calculation.

Codepython
import math

def conv_output_size(n, f, p, s):
    # n=input, f=filter, p=padding, s=stride
    return (n - f + 2 * p) / s + 1

print(conv_output_size(7, 3, 0, 2))  # expected 3

What is the bug in this output-size formula?