Open Source Linux

Apr 16, 2024 · Artificial Intelligence

How Sora’s Text-to-Video Model Is Redefining AI‑Generated Video

Sora, a new text‑to‑video AI model, can create one‑minute videos from textual prompts or static images, delivering industry‑leading fidelity, resolution, and coherent motion by using spatial‑temporal patches inspired by ViViT, and shows emergent capabilities that hint at universal physical simulation.