Google DeepMind releases Genie 4 research preview for real-time interactive worlds with audioMachine translation
Google DeepMind 发布 Genie 4 研究预览，可实时生成带声音的交互世界
Automatically verified and published · Generated and evidence-checked automatically; not reviewed by a human.
Google DeepMind has opened the Genie 4 research preview to trusted testers. Users can generate explorable worlds from text or images, move with keyboard controls, and change scenes with plain-language prompts; ambient sound, music and character dialogue are generated alongside the visuals. Sessions currently last only a few minutes, and objects and visual details can drift after complex prompt edits or longer use.
This source permits summary display only.
Read at the original source