Google has updated Imagen 2, its advanced AI image generator now capable of creating animated clips and editing images, within its Vertex AI platform. This new tool allows users to generate four-second animated videos and alter images using simple text prompts. Imagen 2, which was initially previewed at Google’s I/O conference in May 2023 and later launched in December, marks a significant step forward in the tech giant’s AI capabilities.

The new model builds on the features of its predecessors by not only generating images but also creating short animations with resolutions of 360x640 pixels at 24 frames per second. These animations can portray a range of camera angles and movements, maintaining consistency throughout the sequence. Google highlights that this tool is especially beneficial for organizations in media and advertising, enhancing user engagement through dynamic content.

Additionally, Imagen 2 introduces capabilities like inpainting and outpainting—techniques for refining image content by adding or removing elements and expanding the field of view. These features were previously available in other popular image generators and are now part of Imagen 2’s suite to boost creative possibilities for users.

Google Introduces an AI-Powered Image Generator in Bard
Images generated by Imagen 2 are marked as AI-created through metadata and digital watermarks called SynthID. These images will be appropriately labeled in Google search results.

To address potential misuse, Imagen 2 incorporates safety filters and digital watermarks, known as SynthID, to track AI-generated objects and ensure content integrity. These watermarks are designed to be robust against various modifications, such as compression and color adjustments.