How to check there are how many images after augmentation in the dataset in fast.ai what is the function to know?

Part 1 (2020)
1

function for checking how many images are generated after augmentation,is batch size remain same or become different after augmentation?

2

Generally augmentation should not change the batch size or number of images in the dataset. Maybe if you can provide more context to your question, it will be easier to help.

3

i want to know there are how many images after applying augmentation and there is any function to know how many images are going in model after applying aug_transforms() augmentation in my model.

4

I do not think any such function exists. It is possible to control, what probability the data augmentation is applied which will help you derive an approximate value for your question.

5

okay thanks,so how many images for any unique image are coming by show_batch function and from that,I can count how many images are going in my dataset.i am saying right?please confirm it.

6

Can you try rephrasing the question it is not clear to me.

7

There are no added images via augmentation. Augmentations are stacked on the image when drawn from a batch. Batch to batch these augmented images will be slightly different as randomness is involved in how they’re applied when pulling the batch. If you started with a dataset of 100 images, 100 images will be fed per epoch.

2 Likes
8

It is like that in per epoch different version of augmented image is going?