Digital Humans
Bring avatars to life with generative AI.
Workloads
Generative AI
Industries
Gaming
Financial Services
Healthcare and Life Sciences
Media and Entertainment
Retail / Consumer Packaged Goods
Telecommunications
Robotics
Business Goal
Innovation
Products
NVIDIA ACE
NVIDIA Riva
NVIDIA Nemotron
NVIDIA Audio2Face
-
Overview
-
Technical Implementation
-
FAQ
Digital Humans as AI Agents in Gaming, Enterprise, and Public Sector Settings
Developers, enterprises, and the public sector are exploring how generative AI can accelerate productivity, streamline content pipelines, and provide new and more engaging experiences. From 2D to 3D avatars, whether photorealistic or stylized, creating intelligent digital humans can be challenging—especially in cases where avatar realism is required for human-like interactions and immersion. Capturing intricate human anatomy, facial expressions, and movements, as well as animating characters with natural movements and behaviors, demands sophisticated motion capture systems. Optimizing performance to ensure everything runs in real time is also computationally intensive. To address these challenges, teams are turning to generative AI to find innovative ways of creating interactive digital humans.
NVIDIA ACE is a suite of technologies designed to help developers bring digital humans to life. Several ACE microservices are NVIDIA NIM™ microservices—easy to deploy, highly performant, and optimized to run on NVIDIA Graphics Delivery Network (GDN), a global network of GPUs that deliver low-latency digital human processing to 100 countries, or on NVIDIA RTX™ AI PCs. With individual NIM microservices, platform developers and system integrators can take portions of the ACE technology suite and incorporate them into their end-to-end platform. To streamline the workflow for developers and enterprises alike, NVIDIA's Digital Human Blueprint, powered by ACE technologies and NIM microservices, offers a comprehensive solution to the challenges of creating realistic, interactive digital humans. With an option to use 2D avatars, which are less animation- and computation-intensive, the blueprint provides options depending on the intended use case. Users can interact with a 3D digital avatar from ACE or a 2D avatar from NVIDIA Maxine, based on a customer-service workflow that can connect with people using emotions, humor, and more.
Built on top of NIM microservices, James is a 3D virtual assistant, and Aria is in 2D; both can provide contextually accurate responses.
Interact with James or Aria in real time at ai.nvidia.com.
Quick Links
Harnessing AI for Real-Time Language Understanding, Speech, Animation and Graphics
Built on NVIDIA AI graphics and simulation technologies, NVIDIA ACE encompasses technology for every part of the digital human—from speech and translation to vision and intelligence, realistic animation and behavior, and lifelike appearance.
NVIDIA® Riva
Technologies that enable digital humans to understand human language, translate responses for up to 32 languages, and respond with natural responses.
NVIDIA Nemotron™
A family of large language models (LLMs) and small language models (SLMs) that provide digital humans with intelligence—capable of providing contextually aware responses for humanlike conversations.
NVIDIA Audio2Face™
Technologies that provide digital humans with dynamic facial animations and accurate lip sync. With just an audio input, a 2D or 3D avatar animates realistically.
NVIDIA RTX
A collection of rendering technologies that enable real-time path-traced subsurface scattering to simulate how light penetrates the skin and hair, giving digital humans a more realistic appearance.
Developers can leverage NVIDIA digital human technologies to build their own solutions from the ground up. Or they can use NVIDIA’s suite of domain-specific AI workflows for next-generation interactive avatars for customer service, humanoid robots for virtual factories, digital experiences in virtual presence applications, or AI non-player characters (NPCs) in gaming. Generative AI models can be both compute- and memory-intensive, and running both AI and graphics on the local system requires a powerful GPU with dedicated AI hardware. ACE is flexible, in allowing models to be run across cloud and PC, depending on local GPU capabilities to ensure the user gets the best experience.
"The combination of NVIDIA ACE microservices and the Inworld AI Engine enables developers to create digital characters that can drive dynamic narratives, opening new possibilities for how gamers can decipher, deduce and play."
Kylan Gibbs, CEO, Inworld AI
“Generative AI-powered characters in virtual worlds unlock various use cases and experiences that were previously impossible … Convai is leveraging NVIDIA ACE technologies such as Riva automatic speech recognition and Audio2Face to enable lifelike non-playable characters with low latency response times and high fidelity natural animation.”
Purnendu Mukherjee, Founder and CEO, Convai
News
Get Started
Top game and digital human developers are pioneering ways ACE and generative AI technologies can be used to transform interactions between players and NPCs in games and applications. Learn more about NVIDIA ACE, future developments, and early access programs.