image: This single frame pose compares a real actor with his digital counterpart in SwRI’s Biomechanical Evaluation and Animation Motion Capture (BEAMoCap™) system. The markerless motion capture tool converts video into realistic 3D animations without requiring the conventional marker suits worn by actors.
Credit: Southwest Research Institute
SAN ANTONIO — April 8, 2025 – Southwest Research Institute has launched a new markerless motion capture system that simplifies how film and gaming studios capture human movement for 3D animations. SwRI’s Biomechanical Evaluation and Animation Motion Capture (BEAMoCap™) tool converts video into realistic 3D animations without the conventional marker suits worn by actors.
BEAMoCap won a 2025 Technology Innovation Award from the National Association of Broadcasters (NAB). SwRI staff accepted the award at the NAB Show at the Las Vegas Convention Center, where attendees can visit Booth No. W3843 to see a live demonstration through April 9.
“This mocap system captures and translates actor movements into digital animations with unmatched speed and accuracy, ensuring a higher level of detail and realism in animated characters,” said Jonathan Esquivel, a computer scientist in SwRI’s Intelligent Systems Division who led software development. “This benefit is paramount for game developers striving to create lifelike experiences and filmmakers seeking to push the boundaries of animated storytelling.”
Conventional motion capture, or mocap, requires actors and athletes to wear body suits with dozens of infrared markers, a costly and laborious process. BEAMoCap is a camera-based solution that eliminates marker suits by using artificial intelligence and machine vision algorithms to predict kinetic movement across dozens of joints and related body parts.
BEAMoCap optimizes technology previously developed by SwRI for sports science and human performance. It fuses the precise joint prediction models used for biomechanical analysis with 3D models, meshes and skeletal rigging used for the creation of animated characters.
Movements captured using conventional mocap techniques require significant correction and cleanup. BEAMoCap gives animators more accurate movement based on kinematic modeling designed by SwRI’s biomechanical engineers. Researchers also developed a data cleaning solution to minimize artifacts that requires less modification than raw data from typical marker systems.
“The key benefit of BEAMoCap for the animation community lies in its ability to drastically reduce the production timeline,” Esquivel said. “Game developers and film animators can achieve more dynamic and responsive character animations, allowing for faster turnaround times and greater creative flexibility.”
BEAMoCap leverages SwRI’s Engine for Automatic Biomechanical Evaluation (ENABLE™), a system used by professional and collegiate sports programs to analyze biomechanics and optimize physical performance. The technology’s development is documented in published research referenced at https://enable.swri.org, which includes a white paper explaining algorithm training.
“We are excited to apply our neural network-based technology to animation, film production, digital twins, sport science and other industries where motion capture data is needed,” said Kisa D. Bordas, an SwRI computer scientist who worked on the project.
The BEAMoCap research integrates the ENABLE algorithms into a motion capture-to-animation workflow. Researchers developed a procedure to record motion and process output. Steps include importing motion data, configuring digital actors and attaching motion to digital actor joints. To remove barriers to adoption, BEAMoCap is compatible with MotionBuilder, Autodesk Filebox (FBX) and other animation formats for easy integration into existing workflows.
The team has several ongoing research projects to enhance BEAMoCap and related technology.
To learn more, watch a video demonstration or visit SwRI’s https://www.swri.org/markets/biomedical-health/biomedical-devices/biomechanics-human-performance/markerless-motion-capture-laboratory page.