[AI EXPO KOREA 2026] silentlab Proposes 'Barrier-Free Information Access' Structure with Motion Capture-Based 3D Virtual Sign Language Technology

[산업일보]
silentlab, a Taiwanese AI sign language technology company, participated in AI EXPO KOREA 2026, held at COEX in Seoul from May 6 to 8, to unveil its 3D virtual sign language solution.
silentlab is an enterprise that digitalizes sign language services by combining AI with 3D virtual characters. RuiChen, the representative of silentlab, stated, "Sign language is not merely a set of hand gestures but a language with its own independent grammatical system. For those born deaf, written text can feel like a foreign language; therefore, the accuracy of information delivery through sign language is higher than text, especially for critical information such as emergency evacuation routes."
The core technology is a "virtual sign language production pipeline" that records the movements of sign language interpreters using motion capture and then converts them into 3D animations by applying a proprietary correction algorithm. As sign language is a multimodal language where hand gestures, facial expressions, mouth shapes, and upper body movements must function simultaneously to complete a meaning, silentlab explained that they adopted a method to process these elements integrally.
"Mapping raw motion capture data directly can lead to joint errors or unnatural movements," RuiChen explained. "We ensured accuracy through our independent correction technology and established a structure to process and synthesize various contents by recombining movements based on our accumulated 3D sign language database."
A representative output of this technology is the virtual sign language ambassador, 'Xiaolü'. According to silentlab, a pilot virtual sign language service was operated at a Taiwanese professional baseball stadium in November 2025, and in January 2026, the character appeared in a performance format bridging physical and virtual spaces at a New Year's event in Taoyuan.
The business structure is a B2B consulting model that implements virtual sign language services for corporations and public institutions. The company added that they provide a one-stop response from customized character creation to content planning and on-site application, with all content undergoing accuracy inspections in collaboration with sign language experts in Taiwan. The application media covers 2D and 3D video, as well as VR, AR, and MR, and is structured to be utilized for accessibility improvement projects linked to corporate ESG strategies.
RuiChen expressed, "Through this participation, we intend to explore possibilities for cooperation with companies and institutions within South Korea."