SignAvatar is an AI sign language translator for public announcements. It is described as a way to translate airport and other transport-hub announcements into sign language and 30+ languages, with a focus on helping international passengers and Deaf or hard-of-hearing people understand announcements in their native language visually. The site also names innovation and PAX officers, as well as accessibility directors in the aviation industry, as intended users.
The system translates speech in real time into text in 30+ languages, American Sign Language, and International Sign Language. It is presented as a way to let passengers, staff, executives, and media notice that an airport is doing accessibility work, and it is framed as a software layer on top of existing public address software. According to the page, it receives audio feeds from any PA system through standard audio interfaces, does not require modifications to current PA hardware, and can display announcements on FIDS/GIDS screens, Wi‑Fi captive portals, browser pages within airport or airline apps, or any device that supports a browser. Passengers can also access it through a QR code or URL, and the page says no special software, user registration, or personal details are needed. It states that translations appear visually within 3 to 4 seconds of the original announcement.
The page also says the typical workflow is about six weeks. It mentions possible regional sign-language rollouts, including LSQ and Libras, but gives those as options rather than as a settled supported list.
SignAvatar is an Other translation project. It enables public announcements to be instantly translated into sign language and multiple languages, improving accessibility in transport hubs. SignAvatar is a B2B product aimed at transport operators and accessibility managers. It runs on the web and the command line.
SignAvatar first shipped in 2025. Key capabilities include speech-to-sign translation, multilingual support, and real-time announcements.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do