Google taught its AI to turn a text description into a high-quality video

The competition between the Search giant and OpenAI is growing. Google is taking a new step in the field of video generation using artificial intelligence. DiscussGoogle taught its AI to turn a text description into a high-quality video© YouTube

The company introduced Imagen 2 – a series of models on the Vertex platform AI that can create short videos based on text descriptions.

The Imagen 2 model has advanced capabilities, including image editing and augmentation.

The most important update is the text-to-text feature “live images” (“text to live images”), which allows users to create videos from text queries.

It is worth considering that the videos generated by Imagen 2 are currently short four-second “live images”. Detailed customization options and video resolution may be lower than existing tools.

Google is rolling out SynthID technology to add cryptographic watermarks to videos, which should reduce the risk of deepfakes.

However, there are still questions about data sources used to train AI and potential threats of copyright infringement.


Date:

by