Google Releases AI Video Model "Veo"
Google has introduced Veo, a generative AI video model now available in private preview via Google's Vertex AI platform. Unveiled in May 2024, Veo enables businesses to create high-quality 1080p videos from text or image prompts, offering various visual and cinematic styles. While the preview doesn't specify length restrictions, initial announcements indicated clips could exceed a minute.
An example showcases Veo's ability to maintain consistency in a dog's fur pattern and collar throughout its movement. Additionally, the latest version of Google's Imagen 3 text-to-image generator is set to be available to all Google Cloud customers, featuring prompt-based photo editing and brand integration into images. Both Veo and Imagen 3 incorporate safeguards against generating harmful content and utilize DeepMind's SynthID technology to embed digital watermarks, aiming to reduce misinformation and misattribution.
With Veo's release, Google positions itself ahead of competitors like OpenAI, whose similar product, Sora, has yet to launch. Companies already employing generative AI report increased revenues, highlighting the growing significance of such technologies in content creation.
