Hi, thanks for your interest. We only polish and offer the code and the model for DeepMIH with 2 images hidden, which contains the main experiments in the paper. For DeepMIH-P, it was added in the Response process, and we only added a commonly used vgg_loss, which can be found in lots of low-level works. Specifically, we mainly referred to [HiDDeN](https://github.com/ando-khachatryan/HiDDeN).

[excellent666]

Hi, thanks for your interest. We only polish and offer the code and the model for DeepMIH with 2 images hidden, which contains the main experiments in the paper. For DeepMIH-P, it was added in the Response process, and we only added a commonly used vgg_loss, which can be found in lots of low-level works. Specifically, we mainly referred to HiDDeN.

Originally posted by @TomTomTommi in #1 (comment)