Search Unity

Automate facial rigging with cloud-based machine learning in Ziva Face Trainer

July 31, 2023 in Engine & platform | 4 min. read
Still captures of a Ziva Face Trainer rig showcasing extreme facial expressions
Still captures of a Ziva Face Trainer rig showcasing extreme facial expressions

Is this article helpful for you?

Thank you for your feedback!

Ziva Face Trainer’s automatic facial rigging presents a new and transformative solution for character animation using cloud-based machine learning (ML).

By combining machine learning deformations, automatic face rig generation, and cloud accessibility, Ziva Face Trainer simplifies the facial rigging process. This enables you to focus on creativity while delivering high-quality animatable puppets.

With its time-saving features, real-time performance, and versatility, Ziva Face Trainer promises to be an invaluable tool for studios seeking to streamline the character animation pipeline and achieve high-quality animatable puppets capable of expressing a vast array of facial expressions and poses.

A Ziva Face Trainer rig showcasing extreme facial expressions

Efficient ML-powered rigging

Ziva Face Trainer leverages a vast library of 4D scan data and ML algorithms to transfer performance onto any humanlike character. As an artist, you can easily prepare your static face mesh to match Ziva Face Trainer’s topology and upload it to the cloud-based service.

In around an hour, Ziva Face Trainer can generate a film-quality facial rig suitable for linear content or real-time 3D animation. This drastically reduces the time required for rigging.

Time and resource savings

Traditionally, creating a high-level facial rig would take weeks or even months, even for expert character technical directors and artists. With Ziva Face Trainer, the process is streamlined, and a fully rigged face can be obtained in a day. This saved time not only enhances efficiency but also reduces the need for extensive headcount or outsourcing, making it an accessible solution for studios of all sizes.

Real-time facial manipulation points

Cloud-based accessibility

Ziva Face Trainer’s cloud-hosted nature ensures easy web access at all skill levels. By leveraging cloud infrastructure, the tool becomes readily available, eliminating the need for complex setups or hardware investments. You can seamlessly access and use Ziva Face Trainer from anywhere, promoting collaboration and flexibility in character animation workflows.

High performance and real-time playback

Ziva Face Trainer’s generated face rigs are very efficient, with a runtime size of only 30 MB and real-time frame rates of 3 ms per frame on a single CPU thread. The tool’s integration with the Ziva RT player enables seamless playback within chosen digital content creation applications or game engines, allowing the team to focus on artistic aspects without compromising on performance or visual quality.

Innovation and versatility

Rather than relying on bone structures, Ziva Face Trainer generates ML-driven parameters. This approach captures the nonlinear interconnectedness of facial movements to enable more realistic and lifelike animations.

Ziva Face Trainer is not limited to humanlike characters. It can also create facial rigs for humanoids and creatures for more diverse storytelling and gaming contexts.

Discover (for free) how Ziva Face Trainer can accelerate your character creation workflow, and engage your audience with complex and believable performances.

July 31, 2023 in Engine & platform | 4 min. read

Is this article helpful for you?

Thank you for your feedback!

Related Posts