Nvidia has recently made a big splash in the AI space with their text-to-video technology. This new tech allows users to easily create realistic video clips from simple prompts such as “a storm trooper vacuuming on the beach” and “teddy bear playing the electric guitar, high definition 4K”.
Nvidia’s Toronto AI Lab has developed a revolutionary AI called “Latent Diffusion Models” (LDMs) which can generate videos on a low computing power. By building on the work of text-to-image generators such as Stable Diffusion and adding a temporal dimension to the latent space diffusion model, the AI can upgrade still images to high definition videos with a resolution of 1280×2048.
This update will be a huge help to GIF fans who can now have their short 4.7-second clips be much more vibrant and realistic. And with other AI video models from Google, Runway and Adobe, users can now use the models to create 20-second clips and even two-minute-long videos.
These companies are seeing a rapid improvement in the use of AI to generate longer video clips in stock libraries and beyond. With them, the democratization of video creation, as well as automated film adaptations, become more and more doable even with small computing power.
Nvidia is one of the most recognized names in the AI and graphics industry, with a plethora of impressive offerings under its belt. Founded in 1993 by Jensen Huang and Chris Malachowsky, Nvidia quickly became a leader in the GPU market, producing the first GPU Chip in 1999. Now it is a behemoth, with a massive client base and a wide variety of products from AI research and development to graphic cards, automotive, data centers and more. Jensen Huang currently serves as the president and CEO of Nvidia, leading the company to continue expanding in its technology and market-share.
