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

How to make the figure like paper #45

Closed
bh-cai opened this issue Mar 7, 2022 · 6 comments
Closed

How to make the figure like paper #45

bh-cai opened this issue Mar 7, 2022 · 6 comments

Comments

@bh-cai
Copy link

bh-cai commented Mar 7, 2022

Sorry, I am a rookie and I am very interested in your project.I have a question that how can I make the figure like paper?
looking for your reply,thank you very much!

@rpautrat
Copy link
Member

rpautrat commented Mar 8, 2022

Which figure are you talking about?

@bh-cai
Copy link
Author

bh-cai commented Mar 8, 2022

figure 4,figure 5 and figure 12

@rpautrat
Copy link
Member

rpautrat commented Mar 8, 2022

The code for these experiments has not been released. It has not been cleaned and I didn't think that it could have benefited very much the community, so I am not planning to release it.

@bh-cai
Copy link
Author

bh-cai commented Mar 11, 2022

OK,thank you very much,but may I get the code for the experiments of figure 4,figure 5 and figure 12 for you by the eamil ([email protected]) from you? To be sure, if this requeiset compromises your research, I apologize and will not make such a request again.

@rpautrat
Copy link
Member

The uncleaned development code can be found here: #30 (comment)

I don't have time to detail exactly how to re-run each experiment, but you can find the ROC in evaluation/descriptor_evaluation.py, the occlusion experiment in evaluation/occluded_descriptor_evaluation.py, and for the comparison with SuperPoint, you can visualize the correct/incorrect matches with the function 'plot_color_lines' of misc/visualize_util.py. You would need to use the official release of SuperPoint to get the SuperPoint matches (https://github.com/magicleap/SuperPointPretrainedNetwork).

@bh-cai
Copy link
Author

bh-cai commented Mar 14, 2022

OK,thank you very much!

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

2 participants