Not able to Replicate the stated results (binarization, gray clean-up) on DIBCO dataset and noisy office dataset.

[soumyadeepdey]

The model is not converging for the stated datasets till 300 epochs with the training setup mentioned in the repo. Can you point us to the right settings for training the model?

[Ashwani-Dangwal]

Hi, @soumyadeepdey were you able to train the model successfully using the code provided here?

[soumyadeepdey]

No, I was not able to train the model with this code !!!

[Ashwani-Dangwal]

@soumyadeepdey Thanks for the reply. I also implemented the paper you wrote for color rubber stamp removal and it works great but still there are some edge cases that fails to do the task so did you find any model that works for removing rubber stamps?

[soumyadeepdey]

Thanks for reaching out to me, I have not worked with rubber stamp removal for a long time. Not aware of any recent development/ models in this direction.

…

On Wed, Jul 12, 2023 at 12:53 PM Ashwani-Dangwal ***@***.***> wrote: @soumyadeepdey <https://github.com/soumyadeepdey> Thanks for the reply. I also implemented the paper you wrote for color rubber stamp removal and it works great but still there are some edge cases that fails to do the task so did you find any model that works for removing rubber stamps? — Reply to this email directly, view it on GitHub <#4 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADDWKP75E6JP6N3ST5IM4QLXPZGH7ANCNFSM4YQQIMHQ> . You are receiving this because you were mentioned.Message ID: ***@***.***>

-- ------------------------------------------------------------------------------------------------------------------------------------------ Thanks and Regards Soumyadeep Dey mob no : *+919433715948*

[Ashwani-Dangwal]

Okay. But if you have any insights or direction that can be helpful for me then i am all ears. Thankyou!