Alternative rotational invariance and convolution neural networks #171
+799
−178
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Member
BerndDoser
commented
Jun 27, 2024
BerndDoser
commented
- Alternative rotational invariance (proposed by @RomainChazotte): minimal loss of all rotated images and the reconstructed unrotated image
- Improved CNN from https://www.biorxiv.org/content/10.1101/214247v2
- Normalize loss by brightness
- General 2D image data module + Pokemon dataset
- Update plain autoencoder
…rotated images with reconstructed image of not rotated image
- remove one layers - double channels - smaller learning rate - upsampling in decoder
and use no default value
…um over rotated images with reconstructed image of not rotated image" This reverts commit a1c1627.
…nput image and the reconstructed image using the not rotated image.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.