OpenAI’s new “DALL-E2” AI model can generate HD images

Rate this post
New in OpenAI "DALL-E 2" AI models can now edit selected parts of an image to generate an HD image

Earlier last year, we saw researchers at the OpenAI Consortium develop advanced AI models that could generate colorful and artistic images based on specific keywords and text. An AI-based bot called DALL-E was a digital version of a human sketch artist who was able to generate accurate images based on descriptive text. Currently, the OpenAI research team has developed the next iteration of DALL-E with improved functionality. Let’s take a look at the details below.

OpenAI announces DALL-E2 AI model

The DALL-E 2 AI Bot is an improved version of the previous DALL-E model, giving users more access to the tool. The new model offers higher resolution images and lower latency than the previous model, which could only generate 256×256 pixel images. DALL-E 2, on the other hand, can generate 1024 x 1024 pixel images. This is significantly higher than the previous version.

DALL-E2 model Utilize OpenAI’s CLIP image recognition systemm is a computer vision system announced by OpenAI last year and has an image generation function for generating images.New model Users can also select specific parts of an existing image and instruct the bot to edit those parts. Instead of making changes to the entire image. This feature wasn’t available in early models, allowing users to create fairly unique images.

New in OpenAI "DALL-E 2" AI models can now edit selected parts of an image to generate an HD image

Currently, the DALL-E model was available on the OpenAI website, but was never made public to developers or researchers.But many Companies have adopted the concept of DALL-E bots and implemented it in their models... One of the most popular was the WomboDream app, which produces artistic images based on vague explanations.

Regarding the availability of the DALL-E 2 model, unlike the previous model, The DALL-E 2 bot is currently only available to partners and researchers who have been scrutinized for testing purposes. However, this tool has some limitations that prevent users from producing images that contain nudity, hate symbols, and obscene gestures. Researchers and interested testers can now join the waiting list to try DALL-E2 from the official OpenAI website.