Science

Microsoft announces its own DALL-E: surprisingly, it can even create video or process images

DALL-E by OpenAI is not the only artificial intelligence capable of generating images from a short text description. A few weeks ago, Google also unveiled Image, an AI alternative from a company founded by Elon Musk (among others) that Mountain View itself says is capable of producing much more realistic and high-quality designs. Today Microsoft has joined the competition. It does this with the help of NUWA-Infinity AI, which can not only create images from texts, but also convert a static drawing into a video.

Microsoft describes NUWA as “a multi-modal generative model designed to generate high-quality images and videos from given text, image, or video.” Therefore, its operation is not much different from what DALL-E or even Image (Google) can do. However, it has a number of advantages over the two artificial intelligence models. This is the only AI capable of generating a video from an image generated by a text description. In addition, AI can also generate videos directly from the description.

Compared with DALL-E, Imagen and Parti, NUWA-Infinity can generate high-resolution images of arbitrary size, and also supports the creation of long videos.

NUWA, Microsoft’s artificial intelligence, can also extend any type of image.

NUWA, Microsoft’s artificial intelligence that generates images and videos from a text description, is also capable of… “stretching” any image and creating a larger, higher resolution image. Artificial intelligence, in particular, captures the information contained in the original photo and, depending on its parameters, generates another, much more complete one. NUWA, for example, could expand on Vincent van Gogh’s Starry Night. What’s more, it does so with identical details to those presented in the original design and a very accurate continuation.

At the moment, Microsoft has not provided more details about NUWA, other than a few examples that show the potential of this AI and how it is able to convert text to image, image to video, or text to video. expand any design. This is certainly an interesting alternative to DALL-E and Imagen, although these two algorithms have their own advantages.

Image, for example, generates much more realistic drawings, although it is not yet available to users. DALL-E, on the other hand, offers less realistic images but is more accessible to users as it is available through a public beta, albeit with limited access.

Back to top button

Adblock Detected

Please consider supporting us by disabling your ad blocker.