У нас вы можете посмотреть бесплатно UPSCALE any image for FREE with AI | Stable Diffusion или скачать в максимальном доступном качестве, видео которое было загружено на ютуб. Для загрузки выберите вариант из формы ниже:
Если кнопки скачивания не
загрузились
НАЖМИТЕ ЗДЕСЬ или обновите страницу
Если возникают проблемы со скачиванием видео, пожалуйста напишите в поддержку по адресу внизу
страницы.
Спасибо за использование сервиса ClipSaver.ru
We will see HOW I use AI to upscale images up to 4k or more. And a little bit on what to do with the upscaling errors. (Remember that with youtube's compression you can't see the details as much as I can, but it is what it is) I know there probably are better options out there, but I hope this video helps still. Feel free to discus what you think in the comments! If you want the extension let me know in the comments and I'll try to fix it a little more to publish it. Btw I just realized that I don't really explain how to fix the skin all that much xD Here are some tutorials on how to do that better: Removing shine by PiXimperfect: • Remove UGLY Skin Shine - EASY Photoshop Fix! Frequency separation by Ronnix Photography: • High-End Skin Retouching Beginner Photosho... Join our Discord: / discord ----------- Links used in the VIDEO -------- Download the Ultimate SD Upscale: https://github.com/Coyote-A/ultimate-... Download main CN Models: https://huggingface.co/lllyasviel/Con... Download CNXL models: https://civitai.com/models/136070/con... Download Controlnet for Illustrious models: https://civitai.com/models/929685/noo... ----------- Useful links -------- Afroman4Peace workflow on civitAI: https://tensor.art/workflows/77227780... I honestly havent tested any other than that xD But on civit AI there are a bunch of upscaling workflows for Comfy_UI. Most of them work in similar ways to what we do in the video, just directly. Older but useful info (by @darkstoorm on our discord -goated upscaling master-): https://gist.github.com/DarkStoorM/4b... ----------- Social Media -------- -Instagram: / not4talent_ai -Twitter: / not4talent ------------ Explanation for AI beginners ------------- I'm using Stable Diffusion, a free, opensource, uncensored gen-AI for images. More specifically the Automatic1111 UI, and Forge UI whenever I need to use FLUX (You can use just Forge UI, since it's basically an upgraded version of A1111). Stable Diffusion is the base AI, and it has 3+ variations that people use. -Sd 1.5 (the oldest version, low prompt comprehension and generates at low resolutions, but has the largest amount of extensions and trained LORAs) . -SDXL (the second oldest version. The base is decent but people have re-trained the checkpoints in a way that vastly improve the quality of it. Mainly for anime generations. Though these new models have some specifics to them. Like the need of their own controlnet models, or LORAS. Since they are not compatible with the base SDXL models). The models I use in the video are mainly these. One is an Illustrious based checkpoint. And another (the one needed to use the LORA) is a PonyXL based checkpoint. -Flux (The newest version for Stable Diffusion. Great prompt comprehension and resolution. Probably the best so far when it comes to just text to image. Though there is a current lack of LORAs and controlnet models for it. Which makes me not want to use it all that much except for specific cases). I explain Controlnet and Image2Image more in these 2 videos: Image2Image Explained: • STOP making BORING AI art Controlnet Explained: • Get FULL creative control over Stable Diff... If you want to use StableDiffusion locally like I do, you'll need a decent PC. How to install forge and some explanations by @pixaroma • How to Install Forge UI & FLUX Models: The... Make sure to subscribe if you want to learn about AI and Art. Grow with the community as we surf the AI wave :3 0:00 Intro 0:20 Beginner mistakes when upscaling 1:05 Most basic of upscales 1:35 How upscale works 5:38 Upscaling methods 10:55 Enhancing details (Darkstoorm inspired xd) 12:30 I tried making an extension for it 13:40 For anime and real photos? 14:30 summary of what the method is 15:00 fixing Upscale-related errors #aiairt #digitalart #automatic1111#stablediffusion #ai #free #tutorial #prompting #sd #digitalart #control #controlnet #digitalart #artist #upscale #upscaling #4k #AIupscale