OpenAI Unveils DALL-E 2 With Advanced Image Editing Capabilities
OpenAI launches DALL-E 2, an upgraded text-to-image AI model that features higher resolution and new tools for editing existing photos. The system builds on the original model by allowing users to add, remove, or blend elements within images using simple text prompts.
OpenAI releases DALL-E 2, an upgraded version of its text-to-image generation model that delivers higher resolution and lower latency than the original system. Like its predecessor, this new AI creates fantastic pictures from written descriptions, but it now incorporates powerful new capabilities for users to manipulate existing visual content.
The updated model introduces an "inpainting" feature that allows users to select a specific part of an existing picture and instruct the AI to edit it. This means a person can add a glass of water to a table or remove an object while the system intelligently adjusts related details like shadows to maintain a realistic appearance.
DALL-E 2 also provides a variations feature that generates alternative versions of an uploaded image and enables the blending of two different pictures into a single composition. Built on OpenAI's CLIP computer vision system, this advanced generator is currently available for testing by researchers and shows strong potential for future third-party application integration.