Yapay Zekâ & Teknoloji · Goril Radar’da yayın:

Google Genie 4 adds audio and real-time prompting to interactive worlds

Kaynak: TechCrunch

Kaynak yayın tarihi:

Ses dalgaları içinden geçen kâğıt çerçeveli minyatür dünya; gerçek model çıktısı değil
Ajans Goril — özgün kapak görseli

TechCrunch reports that Google DeepMind has unveiled its Genie 4 world model as a research preview. The report outlines features such as audio matched to the generated environment and the ability to alter weather, camera angles, people, or objects using text prompts while the simulation is running.

The up to two minutes of scene consistency and physics improvements cited by the source are claims made within the scope of the announcement. These do not imply independent benchmark testing or unrestricted public access. In terms of design, the development highlights a transition from generating static visuals to crafting an environment that users can actively manipulate.

TechCrunch: Orijinal haberi oku

Bu metin, kaynak haberdeki doğrulanabilir bilgilere dayanılarak hazırlanan özgün bir editoryal özettir.

← Goril Radar