[Experiment] Plan A: Configurable inference resolution for P2 model validation#362
Draft
xiaotianlou wants to merge 2 commits intomainfrom
Draft
[Experiment] Plan A: Configurable inference resolution for P2 model validation#362xiaotianlou wants to merge 2 commits intomainfrom
xiaotianlou wants to merge 2 commits intomainfrom
Conversation
Allows testing P2 models at higher resolutions (e.g. 736x960, 816x960) to evaluate if increased resolution recovers P2 small-target advantage.
Previous implementation scaled both W and H, producing 1305x734 at scale=0.68 which exceeds Orin memory. Now width stays at WIDTH/2=960, only height scales: 0.68 -> 736x960, 0.76 -> 816x960.
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Adds
--scale-factorparameter toconvert_tensorrt.pyandaccuracy_benchmark.pyto test P2 model at higher deployment resolutions.Changes
src/benchmark/convert_tensorrt.py: Added--scale-factorarg, configurable height scaling with fixed width=960src/benchmark/accuracy_benchmark.py: Passes scale_factor to engine conversionResults
Conclusion
Increasing resolution does NOT recover P2 accuracy. FP16 quantization loss on 43K anchors dominates. Plan A is not viable for deployment improvement.