NVIDIA collaborated with Convai to develop the demo and showcase the capabilities of ACE. The technology can be deployed both in the cloud and locally, utilizing NVIDIA hardware. ACE leverages NVIDIA NeMo to construct, customize, and deploy large language models that can be tailored with specific lore and character backstories. The system incorporates guardrails to ensure appropriate conversations and utilizes Riva, a speech recognition and speech-to-text tool, along with NVIDIA’s Omniverse Audio2Face, which creates dynamic facial animations for game characters to match speech tracks seamlessly.
The Kairos demo was built on Unreal Engine 5, highlighting NVIDIA’s advanced ray-tracing capabilities and other GPU features. While the visuals of the demo were visually captivating, it was evident that the dialogue generated by the AI system could be significantly improved. NVIDIA has not yet announced specific games that will incorporate ACE, but it revealed that Stalker 2: Heart of Chernobyl and Fort Solis will employ Omniverse Audio2Face.
The introduction of ACE marks an exciting advancement in gaming technology, promising a more immersive and interactive gaming experience. By enabling natural conversations with NPCs, ACE has the potential to enhance storytelling, character development, and overall gameplay. While the demo showcased the initial capabilities of ACE, there is considerable room for improvement to achieve more realistic and dynamic dialogue exchanges in future game titles.
