Calculating FID Dose not require image size as 299x299?
mathfinder opened this issue · comments
Defa Zhu commented
The InceptionV3 is trained with image size as 299x299.
But I don't find the code that resize image as 299x299. Dose calculating FID not require this
process? Are the images resized offline?
HRamses commented
Hey! If you open the model graph in tensorboard, you can see that the input is resized to 299x299 by the model.
Tzu-Heng Lin commented
see my explanation here
openai/improved-gan#28 (comment)