DALL-E2 is an impressive demonstration of the power of generative adversarial networks.
Simple text descriptions can be turned into images. It's difficult for a machine learning program to pick up on natural language, but Openai is showing off some images.
DALL-E2 can change the composition, shadows, reflections, and texture of images on the fly. It can combine alien concepts, like generating believable images of an astronauts riding a horse in space or a koala bear dunking a basketball.
Mira Murati, the OpenAI research and product SVP, used DALL-E2 to generate an image from the following text: "macro 35mm film photography of a large family."
Aris used the tool to create pictures of a panda riding a motorcycle in the desert while wearing a red bandana.
Other times I just want to ride in the desert with my panda motorcycle crew. If you see a bunch of pandas wearing red bandanas, that's us! Say hello 🙂
Generated with DALL-E 2. #dalle #openai pic.twitter.com/UdW0U8WSzt
All Rights Reserved © 2024
