Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

GridAnchorRect_TRT get shifted bbox #691

Closed
abx67 opened this issue Jul 15, 2020 · 7 comments
Closed

GridAnchorRect_TRT get shifted bbox #691

abx67 opened this issue Jul 15, 2020 · 7 comments
Labels
Plugins Issues when using TensorRT plugins question Further information is requested

Comments

@abx67
Copy link

abx67 commented Jul 15, 2020

Description

I am trying to try the new commit #679.

But I got the following result:
image

The model supposed to output bbox of traffic light. But it seems to have a small shift for every bbox.

Environment

TensorRT Version: Binary release 7.0; source code 7.0 and chery-pick #679 this commit.
CUDA Version: 10.0
CUDNN Version: 7.0
Operating System + Version: Ubuntu18.04 LTS

@abx67 abx67 changed the title GridAnchorRect_TRT get incorrect output result GridAnchorRect_TRT get shifted bbox Jul 15, 2020
@yshvrdhn
Copy link

I am facing similar issues. @rajeevsrao

@rajeevsrao
Copy link
Collaborator

rajeevsrao commented Jul 16, 2020

I am facing similar issues. @rajeevsrao

Is it with the latest version on master (+7.1 GA release) or for 7.0 GA only?

@yshvrdhn
Copy link

Its from the jetpack for xavier which has the tensorrt version 7.1 I had downloaded the master and built the OSS tools.

@rajeevsrao
Copy link
Collaborator

Can someone please provide a simple repro code?

@rajeevsrao rajeevsrao added bug Plugins Issues when using TensorRT plugins labels Jul 20, 2020
@yshvrdhn
Copy link

yshvrdhn commented Jul 21, 2020

@rajeevsrao Like a sample uff and trt and sample image that shows the predictions ?

@mk-nvidia
Copy link

@yshvrdhn @abx67 Please attach the code / model you're running and the sample image that shows this problem. That way we can run inference using the same model and debug the code. Thanks.

@mk-nvidia mk-nvidia added the question Further information is requested label Oct 27, 2020
@ttyio
Copy link
Collaborator

ttyio commented Jan 12, 2021

I will close this issue since no response for more than 3 weeks, please reopen if you still have question, thanks!

@ttyio ttyio closed this as completed Jan 12, 2021
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Plugins Issues when using TensorRT plugins question Further information is requested
Projects
None yet
Development

No branches or pull requests

5 participants