Skip to content
This repository was archived by the owner on Jul 2, 2021. It is now read-only.

Suggestion to Implement YOLOv2 within chainercv #511

Closed
Paul0M opened this issue Jan 2, 2018 · 4 comments
Closed

Suggestion to Implement YOLOv2 within chainercv #511

Paul0M opened this issue Jan 2, 2018 · 4 comments

Comments

@Paul0M
Copy link

Paul0M commented Jan 2, 2018

Since YOLOv2 is so fast, and Chainer is so dynamic. So an official YOLOv2 in Chainercv will be pretty nice!
Even though, there has been one https://github.com/leetenki/YOLOv2, it's time consuming to train on own dataset(Need to know the details of the author's implementation, especially for the dataset preparation).

@yuyu2172
Copy link
Member

Thank you for your request!

@Paul0M
Copy link
Author

Paul0M commented Jan 16, 2018

New to chainer/chainercv. When I modify leetenki's YOLOv2, I have browsed all the source code of chainer and his repo, but the result doesn't turn out to be good enough. Then I used keras version of YOLOv2, trained on my own data, the result is pretty good. But I prefer chainer/chainercv for it's dynamic and good debuggability. Love chainer!

@machanic
Copy link

YOLO V3 has release, You should re-implement V3 the latest one

@yuyu2172
Copy link
Member

OK. I made another PR that summarizes detection model.
#603

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Projects
None yet
Development

Successfully merging a pull request may close this issue.

3 participants