I am also very interested in DL-based lossy image compression!! I am starting research in it. I believe that there's a lot of scope for DL to improve lossy imag

I am also very interested in DL-based lossy image compression!! I am starting research in it. I believe that there's a lot of scope for DL to improve lossy image compression, since lossy relies on what humans see, and DL is really good at figuring out what humans see. Are you also interested in it? I have found some other people interested in it. Maybe we can start a Discord channel for this?