chainer
chaiNNer
Our great sponsors
chainer | chaiNNer | |
---|---|---|
2 | 46 | |
5,857 | 3,928 | |
0.3% | 5.9% | |
0.0 | 9.8 | |
7 months ago | 6 days ago | |
Python | Python | |
MIT License | GNU General Public License v3.0 only |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
chainer
-
ChaiNNer – Node/Graph based image processing and AI upscaling GUI
There is already an AI framework named Chainer: https://github.com/chainer/chainer
-
Protip: the upscaler matters a lot
Sorry maybe someone could chime in and help but I use chainer to upscale. https://github.com/chainer/chainer
chaiNNer
- ChaiNNer – Node/Graph based image processing and AI upscaling GUI
-
Transmigrations concert visuals remixes
For the video it turned out a bit too "hairy" compared to many of the still images (I believe because of the long landscape aspect ratio), but I ran out of time to fiddle. I used the Seed Travel extension for the animation and ChaiNNer with the 4x-Valar upscaler.
-
How can we, as 3D designers/artists, integrate AI into our work?
While at it, check out ChaiNNer -> https://github.com/chaiNNer-org/chaiNNer
-
What about game assets that target 1080p and you want 4K fidelity?
If you want to do more, there's chaiNNer and CupScale. You need to download an AI model to use those. There are a lot of anime/cartoon models out, so pick one that you like from here. (Note: Upscaly doesn't support these custom models.)
- What workflow is best for upscaling portraits taken by phone camera or DSLR?
-
Are there any google collab scripts or other tools to upscale a bunch of images..?
For local there's cupscale and chainner
-
Selfhosted AI
chaiNNer supports AI models.
-
Civitai should enforce a replicability check
As for the other noticeable increase in quality. Using chainner, I upscaled all my images using an anime focused upscaling model, and then scaled back down to twice their original size. From there I made a chainner workflow that would resize all the images in a directory down to a pixel count I determined, while maintaining whatever ratio the image was in originally. My hypothesis is that if you are training using bucketing, it's best to resize the images to have close to the same pixel count as the resolution you are training at. Supposedly the script does that kind of thing already, but when I did it, and then started increasing the resolution by 64's (576, 640, 704) the detail went up. I had to stop at 704 because that's when I hit the point where my smaller pictures were being increased in size by the script, not decreased in size. I don't know how much this pixel count resizing actually helps, as I haven't tested it extensively yet.
-
Does technology exist to try to convert a low resolution picture into high resolution?
Have you tried chaiNNer ?
-
Finally, Managed to upscale without losing all the micro details..
PS since some people mentioned models like Remacri or others from the upscale wiki, a good application for me to use these is and that u can recommend is chaiNNer. Also i made a page to visually compare these models (thats the goal at least), you can go to the multiple models page and select a model for an example or then you can have a look at my favorites page with just the models that i personally liked best
What are some alternatives?
cupscale - Image Upscaling GUI based on ESRGAN
ultimate-upscale-for-automatic1111
Real-ESRGAN - Real-ESRGAN aims at developing Practical Algorithms for General Image/Video Restoration.
stable-diffusion - Optimized Stable Diffusion modified to run on lower GPU VRAM
stable-diffusion-webui - Stable Diffusion web UI
upscayl - 🆙 Upscayl - Free and Open Source AI Image Upscaler for Linux, MacOS and Windows built with Linux-First philosophy.
SwinIR - SwinIR: Image Restoration Using Swin Transformer (official repository)
VSGAN-tensorrt-docker - Using VapourSynth with super resolution and interpolation models and speeding them up with TensorRT.
A1111-Web-UI-Installer - Complete installer for Automatic1111's infamous Stable Diffusion WebUI
waifu-diffusion - stable diffusion finetuned on weeb stuff
text-generation-webui - A Gradio web UI for Large Language Models. Supports transformers, GPTQ, AWQ, EXL2, llama.cpp (GGUF), Llama models.
HAT - CVPR2023 - Activating More Pixels in Image Super-Resolution Transformer Arxiv - HAT: Hybrid Attention Transformer for Image Restoration