search

JisongXie / StegaStamp_pytorch

StegaStamp of pytorch version
MIT License
56 stars 22 forks source link

Failed to decode #10

Open swyuan521 opened 3 months ago

swyuan521 commented 3 months ago

Hello,

  1. the model I trained has a valid residual image generated by the encoder, but it cannot be successfully decoded. As follows:

Im1552_hidden. png Failed to decode

2.What are the differences between different "borders"? Will it affect the decoder?

Have you ever encountered this situation? Thank you.

JisongXie commented 3 months ago

@swyuan521 Sorry, I haven't worked on this project for a long time.

  1. I think you could debug step by step, and check pixel values of the image.
  2. Now forget. But I think different "borders" means different levels of decoding difficulty, some harder, and some easier. You could do experiments to verify it.
swyuan521 commented 3 months ago

@JisongXie Thank you for your answer, 1.Is it necessary to ensure that pixel values are between 0 and 1 in each step of the code? 2.Will selecting 'no_edge' for borders make it easier for the decoder to decode? Which one did you choose at that time?

ZoeMiyoungChoi commented 1 month ago

Hello :)

When I run decode_image.py, it says that image decoding failed. So I searched and found that bch.decode_inplace(data, ecc) had been changed to bch.decode(data, ecc), but when I used the changed function, it said decode failed.

Thank you.

swyuan521 commented 1 month ago

I think the decoder was not trained well and did not achieve a higher decoding success rate. @ZoeMiyoungChoi