Open AI's Sora has a rival.

The AI ​​war between Google and OpenAI is heating up as both companies continuously release new products. The competition is becoming increasingly intense as Google DeepMind—Google's artificial intelligence research lab—is striving to create a breakthrough in video creation tools.

Google Veo 2 will be a competitor to Sora.

One of DeepMind's most notable new products is Veo 2 – a next-generation AI video creation tool. Veo 2 is the successor to Veo (a simple video creation tool), supporting many products in the Google ecosystem, from YouTube to other applications. It produces high-quality videos, making it easy for users to create content.

Reportedly, Veo 2 is capable of creating clips longer than two minutes, with resolutions up to 4K (4096 x 2160 pixels), delivering video quality four times better than Sora's current resolution. However, this feature is still in the testing phase.

According to DeepMind, Veo 2 will feature a physics-based platform and camera controls, resulting in clearer, more realistic, and more logical video footage. Veo 2 can now more accurately position virtual cameras from various angles. It can also simulate different lenses and cinematic effects, giving videos an artistic, film-like feel. DeepMind has also shared some selected samples, and they look quite impressive with AI-generated video.

Furthermore, there are limitations when users provide data; this model can only generate new content based on specific requests. Although DeepMind hasn't disclosed the exact source of the videos used to train Veo 2, it's likely YouTube, as Google also manages that platform.

Veo 2 will bring several notable improvements: the tool will be able to simulate lenses and cinematic effects; it can capture subtle human expressions, creating more natural and vivid footage. And Veo 2 will continue to be refined before being expanded and deployed to a wider user base.

Source: https://kinhtedothi.vn/sora-cua-open-ai-co-doi-thu.html