YouTube has announced a new update to the Dream Screen for Shorts. The feature is getting support for Google DeepMind's latest video generation model, Veo 2, introduced late last year to rival OpenAI's Sora.
For the unversed, Dream Screen allows users to insert AI-generated images and videos as backgrounds in their YouTube Shorts with just a text prompt. It started out as an experimental feature in June last year and was released to the public in November.
The Veo 2 upgrade speeds up the video generation process, and the model can deliver more detailed and realistic output. The model can generate high-quality videos in a wide range of subjects and styles at resolutions up to 4K and minutes in length. It can understand human movement behavior and real-world physics. You can also specify a style, lens, or cinematic effect while using Dream Screen.
In addition to generating backgrounds for Shorts, Veo 2 can generate standalone video clips that you can add to any of your Shorts. To create a clip, open the Shorts camera > tap Add to open the media picker > tap Create at the top.
Next, describe your idea in a text prompt, then choose an image from the suggestions and pick the desired length of the clip. YouTube uses SynthID watermarks and clear labels to indicate that these videos are generated using AI.
"Together with Dream Screen, Veo 2 generates state-of-the-art, high-quality videos in a wide range of subjects and styles that better match your vision. Plus, we've also made improvements so Dream Screen now generates videos faster than before," the company said in a blog post.
The new features are now rolling out for all users living in the US, Canada, Australia, and New Zealand. YouTube has plans to bring them to other regions in the coming future.
1 Comment - Add comment