Anime4K

I watch AI enhanced movies and/or shows every day when I use Anime4K, a real-time ML upscaler for anime https://github.com/bloc97/Anime4K If you give shit-tier blurry DVD messes, it turns them into nicer blurry messes. If you give it the usual bitrate starved streaming quality show, it looks almost as nice as a blueray. If you give it a blueray... It basically comes out looking exactly the same. I'll usually still add its "darken lines" filter since I'm a fan of strong linework.

> Can you explain what mpv offers over vlc ? People have complained about VLC's decoding quality in the past (no specific source but lots of google results). I'm sure there's valid bugs that have been patched over and some still there, but I'd take it with a grain of salt since VLC is super popular and therefore more noise when it drops the ball. That said, I personally prefer MPV since I can display multiple subtitles at once (for learning foreign languages) and also drop in upscaling / sharpening filters like [1]. Oh and VLC is kinda buggy on my android 14 pixel when streaming from ftp / samba sources (video freezes when seeking too often). [1] https://github.com/bloc97/Anime4K

If you want to avoid manual processing, Anime4K runs in real time as a GLSL shader you can load into MPV (or Plex or something called IINA according to the readme) and still gives great results. https://github.com/bloc97/Anime4K

Players like VLC do much better and let you make your own arbitrary adjustments via GLSL :) In particular, I like the anime4k shader pack, which is ML-based but runs in real time in mpv (and I think VLC as well). While it is tuned for anime (as is obvious from the name), it has decent denoise and deblur which often make YT content more watchable and a restore step that does a really good job with compression artifacts but is a bit too tuned for anime so may not always work, or even make things worse. See https://github.com/bloc97/Anime4K/releases

Personally, it's not really any better than mpv filters like NIS [0] (Nvidia Image Scaling, which does not use AI but is somehow more sharp than these filters, likely because Nvidia used what they learned from AI upscaling to create a better non-AI upscaler; AMD's FidelityFX Super Resolution is also available as an mpv filter on the same page at [0]) as well as Anime4k [1] (which does use AI but only in the training phase to create the initial glsl file which then does not use AI inferencing but plain old image filters). Combine mpv with its yt-dlp [2] and SponsorBlock support [3] and you have a competent way to upscale your own YouTube (and other types of [4]) videos. If you really want to upscale with AI (albeit not at runtime), I've found Topaz Video AI to be the best so far, much sharper than other solutions on the market. [0] https:/

Even digital video created with low-resolution cameras will eventually be viewable in 4K, 8K and beyond because AI upscaling technology is just so great. See Anime4K for examples. https://github.com/bloc97/Anime4K Yes it "guesses" what the missing information is and populates it. Similar approaches will work for live action media too. And successive video frames encode so much extra resolution that a single frame doesn't capture, so there's a lot of information to pull from.

There is also Anime4K [1]. It doesn't use machine learning though. [1]: https://github.com/bloc97/Anime4K

Tangentially related, a real-time video upscaler: https://github.com/bloc97/Anime4K

If I understand it correctly, the whole project is one shader file. Sure it's portable, just pick the glsl file from the repository and plug it into your favorite video player. Edit: uh-huh. https://github.com/bloc97/Anime4K/blob/master/GLSL_Instructi...