Insider's Edit: Google Brings Realism to AI Videos
AI Business
Informing, educating and connecting the global AI community. Next up: #AISummit London, June 2024
Here are the top stories on AI Business this week:
Google's Lumiere: New AI Model that Creates Realistic Videos
Google has unveiled a new text-to-video model capable of generating lifelike videos from short text inputs.
Lumiere creates videos that showcase realistic motion and can even use images and other videos as inputs to improve results.
The model is capable of generating videos comprised of 80 frames. For comparison, Stability’s Stable Video Diffusion clocks in at 14 and 25 frames. The more frames, the smoother the motion of the video.
Lumiere outperforms rival models from the likes of Pika , Meta and Runway across various tests, including zero-shot trials, according to Google's team.
Apple Sold Over 200,000 Vision Pro Headsets in 10 Days
Apple reportedly has sold at least 200,000 of its Vision Pro AR headsets in the first 10 days of the tech giant taking pre-orders.
That is according to a knowledgeable source talking to MacRumors. Apple began taking pre-orders on Jan. 19; Vision Pro will officially go on sale Feb. 2.
At $3,500 for the cheapest model, it is the most expensive AR/VR headset for consumers in the market by far. Its nearest rival is Meta 's top-of-the-line Quest Pro, which costs about $1,000. Meta is the market leader in AR/VR headsets with a 55.2% share, according to research firm IDC. Meta's Quest line of AR/VR headsets starts at $250.
领英推荐
Vision Pro's steep price tag has made several industry pundits wonder if many people would buy them. The 256 GB Vision Pro costs $3,500 but 512 GB will set buyers back $3,700 and the 1 TB version is $3,900.
Meta's 6 Tips to Improve Your Prompts for Llama 2
When using a language model , the right prompt will get you the best results. Software engineers at Meta have compiled a handy guide on how to improve your prompts for Llama 2, its flagship open source model.
The guide provides a general overview of the various Llama 2 models and explains several basic elements related to large language models, such as what tokens are and relevant APIs.
The tips are published under Llama Recipes on the company’s GitHub page, Prompt Engineering with Llama 2 .
Among the tips are employing tactics like Self-Consistency and implementing Retrieval Augmented Generation, or RAG.
In other news
Want more AI news? Subscribe to the AI Business email newsletter.