Skip to content

Conversation

ghost
Copy link

@ghost ghost commented Nov 23, 2017

Current configuration in RandomResizedCrop is fixed between 0.08 and 1.0, which is too much for compact convolutional neural networks such as MobileNet and ShuffleNet. So I make the crop scale and the aspect ratio tunable in RandomSizedCrop to make more flexible augmentation.

@alykhantejani alykhantejani merged commit c4b86b0 into pytorch:master Nov 23, 2017
@alykhantejani
Copy link
Contributor

Thanks!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants