Has Unveiled A New Artificial Intelligence Tool That Can Turn Text Into An Image; A Tool That Can Shape The Future Of Art In Artificial Intelligence With Stunning Results.

This system, called the DALL-E 2, replaces the model unveiled last year. Although the previous version had significant outputs, the new version is a big improvement over the previous one! DALL-E-2 Advanced text comprehension adds four times faster image production and resolution to the technology.

Aditya Ramesh, research fellow penal in this regard to the Next Web, said:

When Randall-E 2, We worked instead of building a more extensive system, focusing on improving image quality and latency.

The new tool also introduces two additional features: the re-interference of existing images and an editing feature called Inpainting.

Inpainting Edits an existing image by analyzing a title in natural language. This feature can add or remove components to an embodiment while integrating desired changes in shadows, reflections, and textures.

DALL-E 2Based on the images and the subtitles associated with them, and the tool helps recognize the relationships between images and words. This process starts with a pattern of random points. After that, the new ideas go through diffusion (Release) and are produced.

The system then gradually converts the pattern into an image when it detects certain aspects of the picture.

Some compositionsDALL-E two, look almost too honest. However, researchers say the system tends to create logical visuals for most of the titles people try out.

For example, I took the above images of an astronaut from a set of 9 produced by this model. Prafulla Dhariwal, A research scientist at OpenAI, Says in this regard that the results are generally stable:

Sometimes, repetition with the model in a feedback loop can be helpful by modifying the command based on a previous interpretation or by using a different style such as “oil painting,” “digital art,” “photo,” or an “emoji” and so on. It can use to achieve a desired style or aesthetic.

In general, the applications of this tool are many. Graphic designers, app developers, media, architects, business illustrators, and product designers can all use this tool for inspiration, new creativity, and editing.

However, besides its benefits, the DALL-E 2 Inherits various biases from its educational data, and its outputs sometimes reinforce social stereotypes.

The team has previously removed explicit content from educational data and included violent, hateful, and adult content in its banned policy.

However, DALL-E may reinforce our understanding of how artificial intelligence sees the world. Open AI hopes such a tool will help them build systems that benefit humanity and are not manipulated to create hatred and deception.