fix #60 - Support for images with multiple objects in Dataset class #62
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
A column called "image_id" is added to the csv files generated using utils.xml_to_csv(). This image_id field is used by core.Dataset's _getitem_(self, idx) function t o identify the index of any specific image. Previously the index an entry inside the csv was used for indexing objects in _getitem_(self, idx) function. However this does not work when each image has more than one object. (See alankbi#60)
Dataset class now supports images with multiple objects. Changes also made to how transformations are applied to ensure scaling and flip transformations apply to bounding boxes of all objects in each image. Modified Model class internal function _convert_to_int_labels to support multiple object labels in same image. Resolves: alankbi#60
Tests updated to ensure compatbility with changes in Dataset class.
|
Awesome, thanks for doing this! I'll take a look at this in the next few days and hopefully we can get it merged in soon. |
My pleasure. Please feel free to reach out with any modification suggestions or clarifications. |
|
Looks great! Merged into master. I still have to release a new version with these changes, but hopefully I'll have that done tomorrow or over the weekend |
|
Great, glad to know! |
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Outdated suggestions cannot be applied.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.
Problem: If a single image contains more than one object, then the dataset class treats each object as a separate item in the dataset. As a result, if there are 'n' objects for any image 'i', then the same image 'i' is repeated in the dataset 'n' times, one time for each object.
Changes:
xml_to_csv()in utils.py was changed to add unique index "image_id" attribute for each image.Dataset.__getitem__(self, idx)in core.py was modified to use "image_id" to index the images. Transformations code also changed to support multiple bounding boxes/labels in the same image.Model._convert_to_int_labels()in core.py changed to support multiple bounding boxes/labels in the same image.Tests: Passing.
test_dataloader()andtest_dataset()to ensure compatibility with changes made.For more information, look at Issue#60.