Feature/sg 1442 sliding window inference for yolonas #1979
Merged
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
BloodAxe
reviewed
May 5, 2024
src/super_gradients/training/models/detection_models/customizable_detector.py
Outdated
Show resolved
Hide resolved
BloodAxe
reviewed
May 5, 2024
src/super_gradients/training/models/detection_models/customizable_detector.py
Outdated
Show resolved
Hide resolved
…ding_window_inference_for_yolonas
…of proccessing for the wrapper
…ding_window_inference_for_yolonas
…ding_window_inference_for_yolonas
BloodAxe
reviewed
May 20, 2024
BloodAxe
reviewed
May 20, 2024
...super_gradients/training/models/detection_models/sliding_window_detection_forward_wrapper.py
Outdated
Show resolved
Hide resolved
BloodAxe
reviewed
May 20, 2024
...super_gradients/training/models/detection_models/sliding_window_detection_forward_wrapper.py
Outdated
Show resolved
Hide resolved
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Outdated suggestions cannot be applied.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.
I am opening this WIP PR as a draft, to discuss some small issues.
Implementing sliding window inference inside CustomizableDetector, it happens that a few arch params are now intersecting with some of the predict() ones.
As @BloodAxe suggested, I run NMS on each tile and finally run NMS (just the IOU part) on the aggregated predictions.
But for predict() we can also pass NMS parameters.
As a temporary solution I just added an IdentityPostPredictionCallback that in the case of sliding window would be used, and internally the user's specified NMS params that were passed in __ init __ would be used for the tiles.
What would be your take on how to handle both predict() specified NMS params and the ones passed in init for the sliding window ?
Is having a different IOU threshold for the final NMS also desired?
@BloodAxe @ofrimasad