DALL-E showed us the last selfie on Earth, now you can type in random prompts to see what a similar AI model comes up with. Called Craiyon, it can draw images from any text prompt, although you do need to be specific to get the desired result(s). For example, you can throw in the keywords “illustration”, “photorealistic”, “high definition”, etc. depending on what you want the image to look like.
For those interested in the technical stuff, the model used is named ‘DALL-E Mini’ and was trained using Google TRC, an initiative by Google to empower researchers to implement and test their models in Cloud Tensor Processing Units (TPUs). Why do some of the faces turn out so weird? It’s not because of the text you inputted, but rather the limitations of Craiyon’s current image encoder. Try it out here.
- 24 daily treats – Behind each door of the LEGO Marvel Studios’ Guardians of the Galaxy Advent Calendar (76231) is a gift to inspire creative...
- Iconic characters – Includes Star-Lord, Rocket, Groot and Mantis minifigures and more to combine with mini builds and accessories
- Mini build models – Buildable play experiences include a drone, the Guardians’ spaceship, a snowman in Thanos’s armor and lots more
While the capabilities of image generation models are impressive, they may also reinforce or exacerbate societal biases. Because the model was trained on unfiltered data from the Internet, it may generate images that contain harmful stereotypes,” said its creator.
