Pre trained model doesn't generalise at all

marcbelmont / cnn-watermark-removal

Fully convolutional deep neural network to remove transparent overlays from images

1.24k stars 228 forks source link

Open kmrabhay opened 5 years ago

kmrabhay commented 5 years ago

Can you please tell me which part might have cause issue of not generalising well. Like which part should I focus on more to improve your code?

marcbelmont commented 5 years ago

The generated watermarks are too simple. One improvement is to generate more complex watermarks to better match watermarks used in the wild.

jjhesk commented 5 years ago

there should be a open repository to share the pretained model and see the results that will help people to improve the model aggressively.