FidelityFX-CLI VS ESRGAN

Hello, I'm having an issue installing FidelityFX-CLI on my computer.

There is no way which works for all the projects. But reading the README or provided documentation is a good start. This one states its a windows solution for upscaling images. And it has precompuked binaries on its releases page.

You can try how the upscaling works yourself: https://github.com/GPUOpen-Effects/FidelityFX-CLI

These were upscaled 2x using FidelityFX-CLI (aka AMD FSR). I think it did a decent job (source res was 720p so 🤷‍♂️ - Better screens could be pulled from an emulator or something). I'm not too familiar with the settings for this tool, so I just used:

For this test I used FidelityFX-CLI for EASU (FSR) and Hybrid for Lanczos. The screenshot is taken from the official FSR demo, but I didn't used the built-in FSR in order to make it a more fair comparison.

You'll need FFmpeg and FidelityFX CLI, but if you don't want to use FFmpeg, any image editing software with Lanczos support should do.

This seems to be based on ESRGAN which is supposed to be higher quality than waifu2x.

I downloaded https://github.com/xinntao/ESRGAN , I run it using command line, no gui.

Nerd info: Uses 3 different algorithms together: https://github.com/CompVis/stable-diffusion For generating source https://github.com/TencentARC/GFPGAN For repairing faces, mouths, eyes https://github.com/xinntao/ESRGAN For upscaling Paint.Net for manual retouching, cleanup and adjustments.

Code for https://arxiv.org/abs/1809.00219 found: https://github.com/xinntao/ESRGAN

Even if it's not a simple black and white image, you can get damn impressive results when upscaling images with GAN-based algorithms, like ESRGAN, together with a suitable pretrained model. Both are free of charge. The software can be tricky to use compared to Photoshop, though.

I use ESRGAN: https://github.com/xinntao/ESRGAN

There are two Generator architectures that are in the code, however these are actually just sitting there unused, likely from work that did not pan out because if you look at their inference code it's not used at all. Instead they directly import the plain old vanilla RRDB architecture from BasicSR yet again another Xinntao repository. Seeing the theme here.