#stablediffusion — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #stablediffusion, aggregated by home.social.
-
🏆 Wan 3.0 just proved open-weight can beat closed models in video generation. It tops the human-preference ranking, ahead of every proprietary rival — and you can download and run it yourself.
-
🏆 Wan 3.0 just proved open-weight can beat closed models in video generation. It tops the human-preference ranking, ahead of every proprietary rival — and you can download and run it yourself.
-
🏆 Wan 3.0 just proved open-weight can beat closed models in video generation. It tops the human-preference ranking, ahead of every proprietary rival — and you can download and run it yourself.
-
🏆 Wan 3.0 just proved open-weight can beat closed models in video generation. It tops the human-preference ranking, ahead of every proprietary rival — and you can download and run it yourself.
-
🏆 Wan 3.0 just proved open-weight can beat closed models in video generation. It tops the human-preference ranking, ahead of every proprietary rival — and you can download and run it yourself.
-
🎨 Where open-source stands in image generation:
#14 Qwen-Image-3.0-Pro — the best model you can actually download
102 ELO points behind GPT Image 2.5 Flare (max) (OpenAI)Blind human preference, not marketing →https://olud.ai/media-leaderboard.html
-
🎨 Where open-source stands in image generation:
#14 Qwen-Image-3.0-Pro — the best model you can actually download
102 ELO points behind GPT Image 2.5 Flare (max) (OpenAI)Blind human preference, not marketing →https://olud.ai/media-leaderboard.html
-
a flying dream for when i go to sleep mad
#Dreamscape #Courtyard #Seascape #Liminal #Pastel #Sky #Clouds #Fantasy #Airy #Flight
#Img2img #AiArt #AiArtists #AiArtCommunity #StableDiffusionin waking life: https://aieris.art/featured/a-flying-dream-for-when-i-go-to-sleep-mad-eris-and-ai.html
-
a flying dream for when i go to sleep mad
#Dreamscape #Courtyard #Seascape #Liminal #Pastel #Sky #Clouds #Fantasy #Airy #Flight
#Img2img #AiArt #AiArtists #AiArtCommunity #StableDiffusionin waking life: https://aieris.art/featured/a-flying-dream-for-when-i-go-to-sleep-mad-eris-and-ai.html
-
a flying dream for when i go to sleep mad
#Dreamscape #Courtyard #Seascape #Liminal #Pastel #Sky #Clouds #Fantasy #Airy #Flight
#Img2img #AiArt #AiArtists #AiArtCommunity #StableDiffusionin waking life: https://aieris.art/featured/a-flying-dream-for-when-i-go-to-sleep-mad-eris-and-ai.html
-
a flying dream for when i go to sleep mad
#Dreamscape #Courtyard #Seascape #Liminal #Pastel #Sky #Clouds #Fantasy #Airy #Flight
#Img2img #AiArt #AiArtists #AiArtCommunity #StableDiffusionin waking life: https://aieris.art/featured/a-flying-dream-for-when-i-go-to-sleep-mad-eris-and-ai.html
-
a flying dream for when i go to sleep mad
#Dreamscape #Courtyard #Seascape #Liminal #Pastel #Sky #Clouds #Fantasy #Airy #Flight
#Img2img #AiArt #AiArtists #AiArtCommunity #StableDiffusionin waking life: https://aieris.art/featured/a-flying-dream-for-when-i-go-to-sleep-mad-eris-and-ai.html
-
fiberscape
Main image is made from the two others - the pattern (from an old mac app), and the close up of denim fibers (because new iphone with very nice cameras.)
The workflow uses the input images plus some minimal prompting to build up the new image. It's an iterative process that constantly checks to see how well the new image fits with both the old ones (metaphorically similar to GAN?)
I get 12 variations from each pair of images. No post to speak of on this one, but sometimes I do a lot and/or combine images.#Abstract #Texture #Fibers #Mesh #Light #Pattern #Soft #Airy
#Img2img #AiArt #AiArtists #AiArtCommunity #StableDiffusionon actual stuff: https://aieris.art/featured/fiberscape-eris-and-ai.html
-
fiberscape
Main image is made from the two others - the pattern (from an old mac app), and the close up of denim fibers (because new iphone with very nice cameras.)
The workflow uses the input images plus some minimal prompting to build up the new image. It's an iterative process that constantly checks to see how well the new image fits with both the old ones (metaphorically similar to GAN?)
I get 12 variations from each pair of images. No post to speak of on this one, but sometimes I do a lot and/or combine images.#Abstract #Texture #Fibers #Mesh #Light #Pattern #Soft #Airy
#Img2img #AiArt #AiArtists #AiArtCommunity #StableDiffusionon actual stuff: https://aieris.art/featured/fiberscape-eris-and-ai.html
-
fiberscape
Main image is made from the two others - the pattern (from an old mac app), and the close up of denim fibers (because new iphone with very nice cameras.)
The workflow uses the input images plus some minimal prompting to build up the new image. It's an iterative process that constantly checks to see how well the new image fits with both the old ones (metaphorically similar to GAN?)
I get 12 variations from each pair of images. No post to speak of on this one, but sometimes I do a lot and/or combine images.#Abstract #Texture #Fibers #Mesh #Light #Pattern #Soft #Airy
#Img2img #AiArt #AiArtists #AiArtCommunity #StableDiffusionon actual stuff: https://aieris.art/featured/fiberscape-eris-and-ai.html
-
fiberscape
Main image is made from the two others - the pattern (from an old mac app), and the close up of denim fibers (because new iphone with very nice cameras.)
The workflow uses the input images plus some minimal prompting to build up the new image. It's an iterative process that constantly checks to see how well the new image fits with both the old ones (metaphorically similar to GAN?)
I get 12 variations from each pair of images. No post to speak of on this one, but sometimes I do a lot and/or combine images.#Abstract #Texture #Fibers #Mesh #Light #Pattern #Soft #Airy
#Img2img #AiArt #AiArtists #AiArtCommunity #StableDiffusionon actual stuff: https://aieris.art/featured/fiberscape-eris-and-ai.html
-
fiberscape
Main image is made from the two others - the pattern (from an old mac app), and the close up of denim fibers (because new iphone with very nice cameras.)
The workflow uses the input images plus some minimal prompting to build up the new image. It's an iterative process that constantly checks to see how well the new image fits with both the old ones (metaphorically similar to GAN?)
I get 12 variations from each pair of images. No post to speak of on this one, but sometimes I do a lot and/or combine images.#Abstract #Texture #Fibers #Mesh #Light #Pattern #Soft #Airy
#Img2img #AiArt #AiArtists #AiArtCommunity #StableDiffusionon actual stuff: https://aieris.art/featured/fiberscape-eris-and-ai.html
-
Генерация видео: Обзор интересных подходов | Text-2-video | Part 2
Освечу базовые концепты из области генерации видео, в этой части разберем уже более современные модели 2025 года, и парочку моделей, с которых все начиналось. Все кратко и четко, только самое основное. Посмотрим на устройство современных топовых SOTA моделей для генерации видео: Wan2.1, Hunyuan video, недавно вышедший подход к облегчению вычислетильных требования FramePack.
https://habr.com/ru/articles/912522/
#computer_vision #нейросети #ml #video_generation #собеседования #stablediffusion #wan21 #comfyui #видеогенератор
-
Генерация видео: Обзор интересных подходов | Text-2-video | Part 2
Освечу базовые концепты из области генерации видео, в этой части разберем уже более современные модели 2025 года, и парочку моделей, с которых все начиналось. Все кратко и четко, только самое основное. Посмотрим на устройство современных топовых SOTA моделей для генерации видео: Wan2.1, Hunyuan video, недавно вышедший подход к облегчению вычислетильных требования FramePack.
https://habr.com/ru/articles/912522/
#computer_vision #нейросети #ml #video_generation #собеседования #stablediffusion #wan21 #comfyui #видеогенератор
-
Генерация видео: Обзор интересных подходов | Text-2-video | Part 2
Освечу базовые концепты из области генерации видео, в этой части разберем уже более современные модели 2025 года, и парочку моделей, с которых все начиналось. Все кратко и четко, только самое основное. Посмотрим на устройство современных топовых SOTA моделей для генерации видео: Wan2.1, Hunyuan video, недавно вышедший подход к облегчению вычислетильных требования FramePack.
https://habr.com/ru/articles/912522/
#computer_vision #нейросети #ml #video_generation #собеседования #stablediffusion #wan21 #comfyui #видеогенератор