OpenAI has released a video detailing how the DALL-E 2 system works. By reading and recognizing text, the DALL-E 2 can combine “concepts, features and styles” to create new artwork. On the company’s website, OpenAI shared an example of an astronaut riding a horse in photorealistic style. Then DALL-E 2 released a photo of just that. However, when switching between styles and settings, you can get a sense of how different the images can be. Our new DALL · E 2 system can create realistic images and art from a description in natural language. See it here: pic.twitter.com/QEh9kWUE8A – OpenAI (@OpenAI) April 6, 2022 Another fascinating example shows that artificial intelligence is changing existing images. Once again, using text and captions, the system can add or remove items from a photo. It can even change shadows, reflections and textures. OpenAI claims that the system has closed the gap between images and text. Using a process called diffusion, DALLE-2 “starts with a random dot pattern and gradually shifts that pattern to an image when it recognizes specific aspects of that image.” As you can tell, DALL-E 2 is the new iteration of the DALL-E OpenAI system created in early 2021. While running with the same basic elements, DALL-E 2 can create and render quadruple images analysis. OpenAI has monitored the capabilities of DALL-E to prevent it from creating hateful images and other abuses. Through the use of DALL-E 2, OpenAI hopes creators will find new ways to express themselves. For professional artists, creators, researchers and developers, OpenAI has a waiting list to access DALL-E 2. Those interested can register here. Image credit: OpenAI Source: OpenAI
