how do I apply my own data to this neural network?

Hi, By applying the formula to calculate the output in 6.3, If we use as input a 224x224 image, after passing through the first conv layer, the output should be of the form 54.5x54.5, if I am not mistaken. However, in the model summary, the output after the first conv is just 54. What does the model do with the 0.5 spared pixel? does it just discard it? is it something the model do automatically?

Thank you for your answer

Hi @le_Minh_Nguyen, great question!

As discussed in 6.3.2, the output value will be the **floor** value of (𝑛−𝑘+𝑝+𝑠)/𝑠