Skip to content

Vectorize box decoding in FCOS model #5247

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.

Sign up for GitHub

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

Jump to bottom
Closed
jdsgomes opened this issue Jan 21, 2022 · 2 comments · Fixed by #6203
Closed

Vectorize box decoding in FCOS model #5247

jdsgomes opened this issue Jan 21, 2022 · 2 comments · Fixed by #6203
Labels
help wanted module: models Perf For performance improvements

Comments

@jdsgomes
Copy link
Contributor

🚀 The feature

The for loop in torchvision/models/detection/fcos.py:80 can probably be vetorized.

        pred_boxes = [
            self.box_coder.decode_single(bbox_regression_per_image, anchors_per_image)
            for anchors_per_image, bbox_regression_per_image in zip(anchors, bbox_regression)
        ]

Motivation, pitch

Simplify/speedup the execution

Alternatives

No response

Additional context

No response
@datumbox datumbox added help wanted module: models Perf For performance improvements labels Jan 21, 2022
@abhi-glitchhg
Copy link
Contributor

Can I try this?

@datumbox
Copy link
Contributor

@abhi-glitchhg absolutely! Thanks for offering to help.

@abhi-glitchhg abhi-glitchhg mentioned this issue Jun 25, 2022
Merged
@abhi-glitchhg abhi-glitchhg mentioned this issue Jul 16, 2022
Merged
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
help wanted module: models Perf For performance improvements
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

Vectorize box decoding in FCOS abhi-glitchhg/vision
3 participants
@datumbox @jdsgomes @abhi-glitchhg