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

Only 10% accuracy for scaled images!!!!! #65

Open
kumarlamichhane opened this issue Mar 2, 2018 · 3 comments
Open

Only 10% accuracy for scaled images!!!!! #65

kumarlamichhane opened this issue Mar 2, 2018 · 3 comments

Comments

@kumarlamichhane
Copy link

CapsNet is said to perform better for scaled images but i trained the network with original images and tested the network with the scaled images to find out the test accuracy to be only 10%... #CapsBoringNet

@GuangyaoZhang
Copy link

CapsNet is said to perform better for scaled images does not mean it can perform well without training. In my opinion, it just has the ability of "finding " it is a scaled picture of original picture, so it is possible if you train capsnet on image with rotation angle 0, pi/2, pi than you test it on pi/3, it will perform well (pi is 3.1415...)

@TarrySingh
Copy link

Kumar — where’s your code? What did you do? 10% accuracy means you’ve potentially also messed up somewhere in your code.

@KumarLamic
Copy link

i trained the network with a training set...
i tested the network with test_data_original to get nearly 99% accuracy..
i rotated the test_data_original with 0.5degree and 1degree
to get another test_data_rotated0p5 and test_data_rotated1
i tested the network against test_data_rotated0p5 and test_data_rotated1 to get the test accuracies around 10%.....
I cloned this repo and feed my datasets..
no clue whats going over here

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

4 participants