{"id":18976,"date":"2022-11-25T19:56:28","date_gmt":"2022-11-25T19:56:28","guid":{"rendered":"https:\/\/mailinvest.blog\/index.php\/2022\/11\/25\/stable-diffusion-2-is-here-but-not-everyones-happy\/"},"modified":"2022-11-25T19:56:28","modified_gmt":"2022-11-25T19:56:28","slug":"stable-diffusion-2-is-here-but-not-everyones-happy","status":"publish","type":"post","link":"https:\/\/mailinvest.blog\/index.php\/2022\/11\/25\/stable-diffusion-2-is-here-but-not-everyones-happy\/","title":{"rendered":"Stable Diffusion 2 Is Here, but Not Everyone\u2019s Happy"},"content":{"rendered":"

<\/a>\n<\/p>\n

\n
\n

\"Tileable<\/p>\n

AI art is 2022\u2019s hottest trend, and it\u2019s all thanks to models like DALL-E and Stable Diffusion<\/a>. Using those, you\u2019re able to generate eerily realistic AI-fueled images. Stable Diffusion 2 has been officially released, bringing several improvements \u2014 and apparently being nerfed in other aspects.<\/p>\n

Stable Diffusion 2\u2019s biggest improvements have been neatly summarized by Stability AI, but basically, you can expect more accurate text prompts<\/a>\u00a0and more realistic images. The text-to-image models are trained with a new text encoder (OpenCLIP) and they\u2019re able to output 512\u00d7512 and 768\u00d7768 images.<\/p>\n

\"\"
Stability AI<\/span><\/figcaption><\/figure>\n

Other models are also improving a lot, including the upscaler, which can now produce much more accurate images, and the depth-to-image model, which can generate new images using both text and an existing image. There\u2019s also an inpainting model that can swap out parts of an image to generate a brand new image.<\/p>\n

However, the new update does have some drawbacks. Users have complained that the new version of Stable Diffusion makes it harder to generate NSFW content as well as art that imitates an actual artist\u2019s style, leading some to claim the new version has been \u201cnerfed.\u201d Given AI art\u2019s heavy criticism for its capabilities to produce real-looking fake images, it wouldn\u2019t be surprising if the model deliberately strays away from producing images that could cause trouble.<\/p>\n

If you want to access the new Stable Diffusion 2, make sure to check out more on GitHub<\/a>.<\/p>\n

Source: Stability AI<\/a>, Engadget<\/a><\/small><\/p>\n<\/div>\n


\n\n<\/div>\n