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.

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

[Feature request] Volumetric Detection Ops #1205

Closed
mibaumgartner opened this issue Nov 9, 2020 · 4 comments
Closed

[Feature request] Volumetric Detection Ops #1205

mibaumgartner opened this issue Nov 9, 2020 · 4 comments
Labels
Contribution wanted Feature request Module: networks network, layers, blocks definitions in PyTorch Module: transform data transforms for preprocessing and postprocessing.

Comments

@mibaumgartner
Copy link
Collaborator

mibaumgartner commented Nov 9, 2020 •
edited
Loading

Is your feature request related to a problem? Please describe.
Add support for volumetric detection ops (NMS and RoiAlign) in Monai.
Especially desired formats of the input and the boxes need to be discussed.

The research working group earlier decided to create a Pull Request in torchvision but it would probably better fit in the Monai repo because it is specifically designed for volumetric/medical data.

pytorch/vision#2402

The Issue is already linked to a PR with an early prototype which needs some rework (CPU implementation, Tests)
Also related to #292 which depends on the box format.

CC
@pfjaeger @wyli
@wyli wyli mentioned this issue Feb 5, 2021
Closed
@wyli wyli added Contribution wanted Module: transform data transforms for preprocessing and postprocessing. Module: networks network, layers, blocks definitions in PyTorch labels May 13, 2021
@dominikj2
Copy link

Hi @mibaumgartner,
Did you manage to find a volumetric RoIAlign op that is open source?

@mibaumgartner
Copy link
Collaborator Author

Hi @dominikj2 ,

there are some implementations out there (MDT, medvision, torchvision PR for the Issue linked above) but (AFAIK) none of them were completely validated for their performance.

Best,
Michael

@dominikj2
Copy link

dominikj2 commented Aug 16, 2021 •
edited
Loading

Hi @mibaumgartner,

Great start. Thanks for the lead. I will investigate and respond here with useful insight.

Cheers,
Dom

@vikashg
Copy link

vikashg commented Jan 5, 2024

closing because of inactivity

@vikashg vikashg closed this as completed Jan 5, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
Contribution wanted Feature request Module: networks network, layers, blocks definitions in PyTorch Module: transform data transforms for preprocessing and postprocessing.
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
5 participants
@wyli @vikashg @mibaumgartner @dominikj2 @Nic-Ma