San Francisco, December 17: Google has launched its state-of-the-art AI video generator, Google Veo 2, and image generation tool, Imagen 3, with powerful capabilities. Google Veo 2 sample videos suggested that the tool would easily take on OpenAI Sora in terms of creating realistic videos from text prompts. Google said Veo 2 could generate videos on a wide range of subjects and styles.

Google announced that its new AI video creator achieved "state-of-the-art" results against the leading models, namely OpenAI's Sora, Runway, Kling, Luma Labs Dream Machines and others. The company said Google Veo 2 would understand "real-world physics and nuances of human movement and expressions" Google's new tool supports both text and image prompts, allowing users to create videos easily with a single click. Another major announcement is the Imagen 3 text-to-image model. Apple Intelligence Controversy: Apple Faces Backlash As Its AI System Sends False News Headline of BBC News About Murder Suspect Luigi Mangione Committing Suicide.

Google Veo 2, Google Imagen 3 Released With State-of-the-Art AI Video and Image Genration

Google Veo 2 AI Video Samples, Watch Here

Google Veo Better Than OpenAI Sora

Google Veo 2 AI Model Key Highlights

Google Veo 2 will support a wide range of prompts , letting creators adjust to their tastes. The tech giant said it could be an "18mm lens" or "shallow depth of field," which will be considered along with other prompts by Google Veo 2. The Sundar Pichai-run company said its new text-to-video model would "hallucinate" less frequently than other models that generate unexpected subjects, extra fingers, and other errors.

Further, the Google Veo 2 AI model will come with a SynthID watermark, which will help identify it as "AI generated." Google is slowly rolling out this model via VideoFX, YouTube, and Vertex AI. Popular tech YouTuber MKBHD praised Veo 2 and said, "If these hand-picked examples are real, they look better than anything I've gotten out of SORA."

Google Imagen 3 AI Model Key Highlights

Imagen 3, another AI tool launched by Google, lets users create images from text prompts. The text-to-image creator also improved in terms of accuracy and quality. The company said," It can now render more diverse art styles with greater accuracy — from photorealism to impressionism, from abstract to anime." The model also achieved state-of-the-art results in side-by-side comparison against leading image generation models. Google said the Imagen 3 model will be rolled out to a global audience through Google Lab's image generation tool, ImageFX, in over 100 countries. iPhone 17 Air Reportedly Enters Mass Production; Know What To Expect From Apple’s Upcoming iPhone Model.

Google Whisk Tool Key Highlights

Google Labs has launched a third tool called "Whisk", which lets the users create images that "convey subject, scene and style you have in mind." After that, the users can bring them together and remix them to create something unique. It can help generate digital plushies, enamel pins, stickers and more. Under the hood, it will combine the Imagen 3 model and Gemini's capabilities to understand visuals and descriptions, writing automatically detailed captions for the images. 

(The above story first appeared on LatestLY on Dec 17, 2024 09:31 AM IST. For more news and updates on politics, world, sports, entertainment and lifestyle, log on to our website latestly.com).