tancik/StegaStamp

Models download link provided.

DMN-sohan opened this issue · 14 comments

Thank you for the author to introduce us to this marvel.
The model link is provided below. Sad to see University of Berkeley to delete the models.

https://drive.google.com/drive/folders/1ZPAKQMYkMHgHHmrdL18WT4XawlEV2hJA?usp=sharing

Thanks for sharing this model ! , can you kindly explain how to use this pre-trained model in the code. Please give command script to run this model. I am new to this domain, so please consider my request. Thanks in advance!

Well, in the project directory, create a folder named saved_models. Place both models in said directory and run the commands provided in README.md according to use case.

Thank you for providing the model, but I get this error.

Detected at node 'save_1/RestoreV2' defined at (most recent call last):
Node: 'save_1/RestoreV2'
saved_models/stegastamp_pretrained/variables/variables.data-00000-of-00001; No such file or directory
         [[{{node save_1/RestoreV2}}]]

Original stack trace for 'save_1/RestoreV2':

I'm not very familiar with tensorflow. Do you have any ideas ?

Thank you for providing the model, but I get this error.

Detected at node 'save_1/RestoreV2' defined at (most recent call last):
Node: 'save_1/RestoreV2'
saved_models/stegastamp_pretrained/variables/variables.data-00000-of-00001; No such file or directory
         [[{{node save_1/RestoreV2}}]]

Original stack trace for 'save_1/RestoreV2':

I'm not very familiar with tensorflow. Do you have any ideas ?

Same error here (missing file). Have you solved it?

NotFoundError (see above for traceback): Restoring from checkpoint failed. This is most likely due to a Variable name or other graph key that is missing from the checkpoint. Please ensure that you have not altered the graph expected based on the checkpoint. Original error:

saved_models/stegastamp_pretrained/variables/variables.data-00000-of-00001; No such file or directory
	[[node save_1/RestoreV2 (defined at encode_image.py:34) ]]

Hello, i have the same question, have you solved it? @Asilvorcarp

`
NotFoundError (see above for traceback): Restoring from checkpoint failed. This is most likely due to a Variable name or other graph key that is missing from the checkpoint. Please ensure that you have not altered the graph expected based on the checkpoint. Original error:

download_models/stegastamp_pretrained/variables/variables.data-00000-of-00001; No such file or directory
[[node save_1/RestoreV2 (defined at encode_image.py:36) ]]`

@swyuan521 I think the file on Google Drive is simply missing or deleted.
However, a new paper shared their replication. You can check it out: https://github.com/umd-huang-lab/WAVES/tree/main/decoders .

thank you very much. but how i use this stega_stamp.onnx? @Asilvorcarp or where should we put the file?

thank you very much. but how i use this stega_stamp.onnx? @Asilvorcarp or where should we put the file?

Have you solved it? I also want to know how to do it. Could you help me?

Sorry, i have not sloved it now. @zkbys I just retrain the model.

Sorry, i have not sloved it now. @zkbys I just retrain the model.

Did you successfully retrain? Can you share the results? I would be very grateful!!!!

I can only run normally, but the watermark effect I get is not as good as in the paper。 @zkbys