DSR and NHK Enterprises exhibited a groundbreaking system for converting 2D video data into 3D animation—jointly developed with our company—at Content Tokyo, held at Tokyo Big Sight on July 2nd.

This system leverages Noema computer vision technology to generate 3D animation from standard DSLR video footage, a process that previously required large-scale motion capture setups. It faithfully reproduces the complex movements of sign language—including hands, body, and facial expressions—with high precision.

The system’s applications extend beyond sign language communication, offering potential uses in disaster prevention systems, integration with various public service platforms, and even in general animation and gaming industries.

At this exhibition, we demonstrated real-time 2D bone data generation from DSLR footage and the advanced sign language representation through avatars. The demonstration received high praise from many attendees.

Moving forward, we aim to achieve even more realistic and refined real-time imagery.

Learn more about the app at https://noema.tech/sign-language-animation/