text-to-video sequences (with camera movements .
In recent months, there have been some exciting ideas in the area of generating different media using artificial intelligence. Google and Meta have now taken it to the next level.
Of course, the technology and the results are not yet perfect, but the rapid development is still exciting to watch.
Google AI creates videos from texts 1
Google Imagen Video: Text to short videos
Google's Imagen project was shown a few months ago. Similar to Dall-E or Stable Diffusion, images can be generated by entering text.
What is new here, however, is that Imagen Video is also ivory-coast number dataset working on text-to-video . This means that short videos can be created from simple text input. Of course, the videos are not perfect yet. But when you consider that this is only created through text input, it is really fascinating.
Higher resolution possible
Resolution and quality are still a problem for both AI images and videos. By using Cascaded Diffusion Models, Google is able to scale the videos up to a resolution of 1280×768.
Source and more videos
On the Imagen Video page, you can find more examples and play around a bit yourself: https://imagen.research.google/video