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

where can I get human_detection.json of MPII? #4

Closed
wangdong128 opened this issue Jan 18, 2019 · 5 comments
Closed

where can I get human_detection.json of MPII? #4

wangdong128 opened this issue Jan 18, 2019 · 5 comments

Comments

@wangdong128
Copy link

I want to test model trained from MPII, where can I get human_detection.json of MPII

@mks0601
Copy link
Owner

mks0601 commented Jan 18, 2019

I implemented codes for MPII dataset, however did not train and test on that. You can convert MPII dataset using tool/mpii2coco.py to convert MPII dataset annotation files to COCO format and use detectron (https://github.com/facebookresearch/Detectron) to detect human. Since detectron uses coco format data, test it on the MPII dataset would be easy.

@wangdong128
Copy link
Author

Thank you for your reply. And how about the human_detection.json of coco2017, can you provide it,thanks

@mks0601
Copy link
Owner

mks0601 commented Jan 20, 2019

It is already uploaded. You can download it in the README.

@mks0601 mks0601 closed this as completed Feb 6, 2019
@Emma0928
Copy link

Thanks for your work. I'm confused about human_detection.json file. Is this file just for human detection? You mean I can seek help from other works for human detection such as https://github.com/facebookresearch/Detectron? If I don't use this step, will it influence the test result?

@mks0601
Copy link
Owner

mks0601 commented May 28, 2019

Of course you can use them. Note that you need to prepare coco 2017 dataset. As far as I know, Detectron has the result of coco 2014. However, as coco 2017 train == (coco 2014 trainval), and coco 2017 val == coco 2014 minival, you can easily convert their result to coco 2017 format.

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

No branches or pull requests

3 participants