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

Preprocessing step in automatic annotation pipeline #1004

Closed
weizhiyi777 opened this issue Dec 25, 2019 · 4 comments · Fixed by #1767
Closed

Preprocessing step in automatic annotation pipeline #1004

weizhiyi777 opened this issue Dec 25, 2019 · 4 comments · Fixed by #1767
Assignees
@nmanovic
Labels
enhancement New feature or request
Milestone
1.1.0-beta

Comments

@weizhiyi777
Copy link

Hi, when I use auto annotation in CVAT, I'm wondering:
The input image is read by opencv, then the channels of it would be in order of BGR. However, the input of my model (converted by openvino) needs to be in order of RGB. How could I tackle this problem? What's more, I don't know how to do some preprocess before inference. Is there any way to do that? It could be convenient when I want to use certain models in auto annotation.

Thanks a lot!
@bsekachev bsekachev added the question Further information is requested label Dec 25, 2019
@bsekachev
Copy link
Member

@weizhiyi777
Hi,

Look at --reverse_input_channels flag in model optimizer

@weizhiyi777
Copy link
Author

@bsekachev
Thanks, I will try!
And if there is any way to add some preprocess operations?

@bsekachev
Copy link
Member

@weizhiyi777
There aren't way in CVAT to preprocess inputs before inference.
But I suppose it's a good point to be implemented.

@nmanovic nmanovic changed the title model needs input image in order of RGB channels when using auto annotation Preprocessing step in automatic annotation pipeline Dec 25, 2019
@nmanovic nmanovic added enhancement New feature or request and removed question Further information is requested labels Dec 25, 2019
@nmanovic nmanovic added this to the Backlog milestone Dec 25, 2019
@nmanovic nmanovic self-assigned this Jul 29, 2020
@nmanovic nmanovic modified the milestones: Backlog, 1.1.0-beta Jul 29, 2020
@nmanovic nmanovic linked a pull request Jul 29, 2020 that will close this issue
DL models as serverless functions #1767
Merged
18 tasks
@nmanovic
Copy link
Contributor

Should be covered by #1767

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@nmanovic nmanovic

Labels
enhancement New feature or request
Projects
None yet
Milestone
1.1.0-beta
Development

Successfully merging a pull request may close this issue.

DL models as serverless functions cvat-ai/cvat
3 participants
@weizhiyi777 @bsekachev @nmanovic