I have a question about your code "DJSCC-for-Wireless-Image-Transmisson"

[ZJU-XWK]

I use your code to train the model successfully. But the MSE(mean square error) is so high. For example, when SNR=20dB and compression ratio= 0.09, the MSE is about 0.009. But in paper named 'Deep Joint Source-Channel Coding for
Wireless Image Transmission", the MSE is about 0.001 when SNR=20dB and compression ratio= 0.09, as fig3 shows in paper. That's a big gap. I want to know that when you train the model, could you reach the performance the same as paper shows?

[irdanish11]

I tried to reproduce the results but I didn't succeed

[irdanish11]

0.2356

[ZJU-XWK]

I have found a problem in your code. In the first layer and second layer of conv layers, you set the padding = 'VALID'. Now I change it to padding = 'SAME' and train the model again. The MSE decreased a lot. You can have a try.

[ZJU-XWK]

[irdanish11]

@xwk111 can you send me a pull request so that I can merge the update code with master branch, and also before doing that please add me as a reviewer so that I can review the changes, thanks

[ZJU-XWK]

I want to ask you a question. How to reproduce the image transmission system(JPEG + LDPC + QAM)? Are there any python libs for JPEG algorithm with variable bit per pixel?