SYSTEM OVERVIEW

YoomTrack Performance Capture leverages cutting-edge computer vision algorithms, an array of strategically placed cameras, and sophisticated machine learning methodologies to precisely track human movement in three-dimensional space—completely eliminating the need for physical markers on the subject. This technology enables the accurate capture and reconstruction of full-body, hand, and facial motion, making it an ideal solution for a wide range of applications, including film production, character animation, virtual performances, immersive virtual reality experiences, video game development, biomechanics research, and sports performance analysis.

FEATURES

No Physical Markers Required

Employs multiple cameras and advanced machine learning algorithms to track movement without the need for markers or special suits.

Multi-Camera Capture Capability

Supports configurations with multiple cameras to capture full-body, hand, and facial motions from various perspectives.

Real-Time Motion Data

Delivers a real-time motion data stream, enabling fast review and efficient selection processes.

High Accuracy and Precision

Offers exceptional accuracy for both simple and complex movements, supported by advanced automated cleanup algorithms.

Natural Performance & Movement Capture

Designed to capture natural, unencumbered performance without requiring traditional marker setups.

Versatile Environment Compatibility

Functions effectively in diverse environments, from controlled studio settings to dynamic, complex locations.

Multi-Angle Reference Cameras

Leverages all available cameras in the setup as potential reference points, simplifying the selection process.

Standardized Data Outputs

Provides output in widely-used formats such as FBX, ensuring seamless integration with 3D animation tools, game engines, and other industry-standard software.

Multi-Actor Tracking

Supports simultaneous tracking of multiple actors, limited only by the available capture space and camera configuration.

Occlusion Handling

Utilizes predictive intelligent tracking to maintain accuracy during partial occlusions, such as when a performer is momentarily obscured by objects or other people.

Fast Calibration

Features fast and automated calibration tools to ensure accuracy is consistently maintained over extended use without holding up production

High-Speed Capture Mode

Designed to handle rapid movements, making it ideal for sports performance analysis, martial arts sequences, and other high-speed applications.

SYSTEM COMPONENTS

Cameras

Type: Vision cameras and optical motion capture cameras.

Resolution: 5 megapixels or higher.

Exposure Time: 1–3 milliseconds.

Number of Cameras: Generally requires 14 to 40 cameras arranged around the capture space, with the exact number determined by the size of the stage and the number of performers

Software

Stage Manager: A streamlined capture stage and production management tool designed to optimize the efficiency of production workflows.

Calibration Tools: Auto-calibration of the system to ensure the accurate placement of captured data within the virtual environment.

Real-Time Data Processing and Review: Provides immediate output of character animations and performance captures directly into digital assets, enabling live review and pre-vis for faster iteration.

Production Manager: A comprehensive data and process management solution that facilitates performance capture workflows, integrates automated cleanup tasks, and enhances team collaboration.

Grid Manager: A centralized resource management system that efficiently allocates processing power to meet production demands.

Solve and Automated Cleanup: Employs advanced algorithms to accurately track body joints and convert raw camera data into fully post-processed skeletons and motion curves.

Integration Support: Offers seamless compatibility with popular 3D software platforms, including Maya, Blender, Unity, and Unreal Engine.

PERFORMANCE SPECIFICATIONS

Tracking Accuracy

Tracking Accuracy: Accuracy Level: Typically within 1–2 cm (0.39–0.79 inches), depending on camera quality and quantity.

Capture Volume

Dimensions: From 3x3m (10x10 feet) for simple setups to 15x15m (50x50 feet) or larger for more complex captures.

Environment: Optimized for studio environments with proper lighting and clear lines of sight; also capable of operating in natural environments given adequate space.

TECHNICAL SPECIFICATIONS

Cameras

Resolution: 5MP or higher

Exposure Time: 1–3 milliseconds

Lens Type: Wide-angle (12mm Computar lenses)

Synchronization: Time-code or genlock support for precise sync

Software and Processing

Data Output: FBX and other standard animation formats

Processing Speed: Real-time tracking with low latency. One minute of cleaning per second of capture in standard configurations (4 Processing agents), system speeds scalable with additional processing agents

Platform Compatibility: Supports Windows and Linux

Integration: Works seamlessly with popular 3D animation tools, game engines, and VFX software (e.g., Maya, 3DS Max), game engines (Unreal Engine, Unity), and visual effects applications (Houdini, Nuke).

CONNECTIVITY

Internet Connection

Options: Offers Ethernet and COEX connectivity for reliable communication between cameras and servers.

SYSTEM REQUIREMENTS

Hardware Requirements (for processing agents):

Processor: Intel® Core™ i9-13900K

Memory: 64 GB (4 x 16 GB) DDR5-4800 UDIMM

Graphics Card: NVIDIA RTX 4090 24GB GDDR6X

Storage: 512 GB HP Z Turbo Drive M.2 SSD

Software Requirements:

Operating System: Windows 10

Additional Software Compatibility: Fully compatible with leading 3D modeling, animation, and game development tools, including Unreal Engine, Unity, Autodesk Maya, and Motion Builder.