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

I want to know heatmap value's range #31

Closed
2p990i9hpral opened this issue Apr 25, 2019 · 7 comments
Closed

I want to know heatmap value's range #31

2p990i9hpral opened this issue Apr 25, 2019 · 7 comments

Comments

@2p990i9hpral
Copy link
Contributor

I want change heat map range to 0~255 for heatmap visualization.

I guessed heatmap range is 0~1 and tried multiply 255, but it was not.
When multiplied by 255, the heatmap's maximum value was about 30.

pleas let me know about heatmap value's range!

@YuliangXiu
Copy link
Owner

Yon can calculate the max-value and min-value from multi channels of heatmap. I think it is 0~1 after sigmoid function

@2p990i9hpral
Copy link
Contributor Author

Yon can calculate the max-value and min-value from multi channels of heatmap. I think it is 0~1 after sigmoid function

unfortunately when I saw heatmap value its range is about 0.00000007 ~ 0.1
I can not set a threshold between joints and not joints.
In other words, I can't classify is person exist in image using heatmap

Thank you for your reply and I am sorry to bother you.
:<

@YuliangXiu
Copy link
Owner

Maybe you should install the latest version of dsntnn. This kind of error has not occurred to me before.

@2p990i9hpral
Copy link
Contributor Author

Maybe you should install the latest version of dsntnn. This kind of error has not occurred to me before.

ok I will try it

@2p990i9hpral
Copy link
Contributor Author

2p990i9hpral commented Apr 26, 2019

Maybe you should install the latest version of dsntnn. This kind of error has not occurred to me before.

I checked the version of my dsntnn. it is already 1.0.1

@YuliangXiu YuliangXiu reopened this Apr 26, 2019
@YuliangXiu
Copy link
Owner

can you visualize some heatmap results to me


plt.imshow(heatmaps[0, 8].data.cpu().numpy())
plt.show()

@YuliangXiu
Copy link
Owner

according to the visualization results, i think you can use a threshold to filter the low-score keypoint

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