Real-ESRGAN Tutorial

My main channel where I introduce the latest fascinating AI tools

Main video

My copy of Real-ESRGAN GitHub
Join my discord if you have any questions

[Quantum_Nebula]

This is wonderful, I will be upscaling the rest of Early Edition! Woooo! This is perfect timing.

[nefwaenre]

Just used the collab gen, my Zod~!! It's fantastic!!! Thanks so much, my dude!!

[mentechiara]

Thank you so much I’m making soft so fast now back then I was confused, thanks

[travelthetropics]

Great tutorial! How do you compare the speed between Vulkun & CUDA versions? specially for video ?

[tehprobot]

Awesome work & great guide! I have one question & one pointer.

Q - Can this be directed to run on all images in a folder structure and output retain folder structure? Ideally for large databases of images.
P - Would be helpful for others to indicate in the beginning that for the Vulkan Install jump here & Cuda Install jump here!

[tehprobot]

Also - For those running into issues with 3070/3080/3090 cards use the following in your Conda env
conda install -c anaconda cudatoolkit
conda install pytorch torchvision torchaudio cudatoolkit=11.1 -c pytorch -c conda-forge

[PrilaRis]

thank you so much! I've been wondering this since the start of your first video 😭❤️
i have several questions if you dont mind :
what does func do?
what does the pre-trained do ? is it for learning similar image/input to make output faster?

[JoaoVictor-fkno]

how to reduce the upscale (for images) from 4x to 2x or 3x?

[ThieuMa-dono]

Thx for this tutorial, it's really helpfull. But I have one big problem : the result video has no audio and the audio of the original doesn't sync with it. I think it's because it's not the same frame rate or some useless frame are removed. Is there a way to fix that?

[ChristianPrince-]

are there any collab version that can enhance bulk bunch of images at once?

[LeBeautiful]

This is dope, thanks for creating such an intuitive tutorial which can transform how we look at older material! Well, better than those "THIS IS 4K ANIME" anyway.

[LachineTokyo]

Thanks for this tuto helped me a lot, its working for me but ... i still have a question, when i download the others models they are not working, the file is in "pth" format does that mean the others models work with python only ?

Also do you know how to use different size of upscale like x1 x2 x3 i tried but it looks like its not working " it work only on animevideomodel x2 & x3 yes but unfortunately no x1 " do you have any information about that thanks a lot in advance

[dna]

why there are warning "vkQUEUE SUBMIT FAILED -4" and then the output image is black only?

[soundchimera]

Using the collab, are we able to upload videos or only images

[FrozzenK]

Getting my master's degree to use this

[Seekah_]

Hey man. How can I make outputs take on the same name as input?

[the_stereo]

Very nice. This process consumed about 1GB RAM, and the NVIDIA GeForce GTX 1050 Ti graphics card warmed up to 61 degrees Celsius. From the 4100x6000 px graphic he made a 12150x24000 px image. Impressive. However, when it comes to complex graphics, it cuts the quality. So the details are flattened.

[wambo]

Is it possible to enhance a whole folder of images automatically?

[JOHN-DUBS]

How do I upload an entire folder of images to the AI

[spaceinvadur]

why am i getting this error? " '.' is not recognized as an internal or external command,
operable program or batch file." im following the steps but cant get shit to work