Let's copy the paper!
"a high-quality oil painting of a psychedelic hamster dragon"...
Hm. Don't think the small model is quite as good...

0 16

Prompt: "mirror of love, by greg rutkowski and james jean" (CLIP guided diffusion, 512x512 openai finetuned on openimages)

10 92

"an illustration of a bunny with a mustache playing a guitar"

Russian on left, OpenAI on right.

1 26

The Prince Reading A Tapestry Of Alchemical Lightning

(CLIP Guided Diffusion [512x512 OpenAI ImageNet])

0 7

Food Storage In An Underground Bunker A, B, C, D

(CLIP Guided Diffusion [512x512 OpenAI ImageNet])

https://t.co/Xre22jdt87

3 45

"a beautiful epic wondrous fantasy painting of [the ocean / lightning / wind / a deep valley]" (OpenAI dVAE + CLIP)

7 56

generated portrait of

That's some spicy meat, ehehehe

0 1

Another generated portrait of

No image references, only text! OpenAI's CLIP model is just so good at compressing all of the web knowledge it's been trained on into only 350MB. Somewhere in there is https://t.co/NfP9wZy94H

0 1

Text-to-image with OpenAI's CLIP is getting more and more surreal. Here is "A banquet hall"

395 2245

Instead of sleeping I've been on openai's Dall e AI and created these godless creatures.
Tag urself I'm the pile of pika-goo!😂

0 1

これで全員デザイナーになれる!

なんてね〜😉

実は面白いニュースを紹介したい!

OpenAIはテキストの元に画像を描き出せるDALL-Eというニューラルネットワークを作った。

例えば、DALL-Eに「アボカドの形の椅子」と指示を出すとこういうコレクションを描き出してくれる。

すごくないですか?😃

6 74

https://t.co/uN8hHApNKg
OpenAIが開発したTransformerベースのtext-to-image(文書から画像を生成)モデル。論文は未発表だが120億のパラメータ(GPT-3の1/10程度)をもつ。単一の物体だけでなく複数の物体も描写可能で、存在しない概念(アボカド椅子)も上手く描写できる。

11 62

全部OpenAIが発表したDALL-EというAIがテキストから画像を生成してる.すごい...
https://t.co/4XNnksU2Qo

170 448

Remember OpenAI's GPT model that was too dangerous for mere mortals? Well, it's now for sale on Azure • The Register https://t.co/BahjGO2I5r

3 4

tried to make some music using openAI's jukebox model. this was supposed to be shakira.
enjoy

3 8

いろんな人に見てほしい 元気が出る動画❣️
もうすぐ18時からYouTubeで公開㊗️🤘

https://t.co/apBQLVbHKD

9 64