[ASK] about the job of document cleanup

[shanyun123456]

Is there the pre-trained model of the job of document cleanup,Would you mind uploading the pre-trained model?
@soumyadeepdey

https://paperswithcode.com/paper/high-resolution-image-synthesis-and-semantic
You have uploaded the implementation of torch in this url,but actually it's the implementation of tf.
if you have the implementation of troch,would you mind uploading it?

[shanyun123456]

@soumyadeepdey

[soumyadeepdey]

I haven't uploaded any code at https://paperswithcode.com/paper/high-resolution-image-synthesis-and-semantic. The implementation is only available in tf as of now. Few of the saved models which I could find is attached with this email. Regards Soumyadeep

…

On Tue, Oct 12, 2021 at 3:16 PM shanyun123456 ***@***.***> wrote: Is there the pre-trained model of the job of document cleanup,Would you mind uploading the pre-trained model? @soumyadeepdey <https://github.com/soumyadeepdey> https://paperswithcode.com/paper/high-resolution-image-synthesis-and-semantic You have uploaded the implementation of torch in this url,but actually it's the implementation of tf. if you have the implementation of troch,would you mind uploading it? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#658>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADDWKP3KZYXQMB4GGZIBJM3UGP7WTANCNFSM5F2D63XA> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>.

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

[shanyun123456]

Sorry about that,i have uploaded a other url.
The right url is https://paperswithcode.com/paper/light-weight-document-image-cleanup-using

[shanyun123456]

Would mind uploading the pre-treained model of M32-color?

[soumyadeepdey]

Sorry for that, I will not be able to share M32-color, since it was trained on proprietary data. Thanks Soumyadeep

…

On Wed, Oct 13, 2021 at 12:51 PM shanyun123456 ***@***.***> wrote: Would mind uploading the pre-treained model of M32-color? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#658 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADDWKP6G5YLBNW7ICLOSPETUGUXNZANCNFSM5F2D63XA> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>.

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

[shanyun123456]

How about the pre-trained model of M32-gary?
I have tested the model in github, but this problem occered,Would mind checking that?

infer code:
#Input image
test_img_name = 'dataset/sample_data/image_56.png'
##########################################################################################
#calling inference engine
out_img_name = 'test_out1.jpeg'
infer_image('checkpoints/M16_noisyoffice.json','checkpoints/M16_noisyoffice_epoch-910.hdf5',test_img_name,out_img_name)

the problem:
NameError: name 'convert2gray' is not defined

thanks

[shanyun123456]

Hi,When i use your code train new model in gpu,it seems always occer the problem of OOM.
How can i change your code to fix that problem?
I set train_batch_size is 1

the problem is:
.....
six.raise_from(core._status_to_exception(e.code, message), None)
File "", line 3, in raise_from
tensorflow.python.framework.errors_impl.ResourceExhaustedError: OOM when allocating tensor with shape[512,512,3,3] and type float on /job:localhost/replica:0/task:0/device:GPU:0 by allocator GPU_0_bfc [Op:Conv2DBackpropFilter]
2021-10-13 15:53:55.915987: W tensorflow/core/kernels/data/generator_dataset_op.cc:103] Error occurred when finalizing GeneratorDataset iterator: Failed precondition: Python interpreter state is not initialized. The process may be terminated.

[sususu1]

hi, do u know how to generate gt data

[soumyadeepdey]

In our case, we generated gt data manually. Other option may be, you can collect few clean pdf images as seed data, and then apply some degradation on it to create noisy images. Thanks & Regards Soumyadeep

…

On Fri, Dec 31, 2021, 7:01 AM sususu1 ***@***.***> wrote: hi, do u know how to generate gt data — Reply to this email directly, view it on GitHub <#658 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADDWKP253SOSVQF4VXLJJSDUTUBXHANCNFSM5F2D63XA> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you were mentioned.Message ID: ***@***.***>