some question about this paper

[Godthumb]

hello, how do you think if i add GAN model into this framework, will I have chance to get more discrimitive feature from this autoencoder?

[MrGiovanni]

Hi @Godthumb

Adding an adversarial loss could be a good idea for higher-quality image restoration, but whether better restoration (more realistic) means better representation is unknown.

This paper (https://link.springer.com/chapter/10.1007/978-3-030-59719-1_24) used adversarial loss in restoring Rubik's cube transformation. Maybe you will find it interesting. But I don't see the ablation study about whether adversarial loss contributes to the transfer learning performance or not.

Thanks,

Zongwei