See-Through Captions

Transparent Display for Real-time Subtitles to Communicate with the Deaf and Hard of Hearing

Real-time transcription using voice recognition is becoming an essential tool to support the daily lives of people who find it difficult to hear. For better communication, it’s necessary to have a system that allows for the use of body language, like facial expressions and gestures, while comfortably preventing misrecognition. We propose a system for real-time transcription using a transparent display. In this system, the results of the transcription are displayed on both sides of the display so that the speaker can check for misrecognition by automatic voice recognition. Additionally, thanks to the high transparency of the display, it’s possible to converse while confirming both body language and text.
This project is one of xDiversity’s initiatives and is being researched at the Digital Nature Group, University of Tsukuba. For inquiries related to the research, please see below.

Inquiries about See-Through Captions

See-Through Captions Project Page

▼ Frequently Asked Questions can be found here:

E-mail: prj_see-through-captionsdigitalnature.slis.tsukuba.ac.jp
(Replace <-at-> with @)
(Contact: Ippei Suzuki)

Authors: Kenta Yamamoto, Kazuhei Suzuki, Akihisa Setsura, Ryosuke Momota, Ryo Iijima, Yoichi Ochiai
Equipment provided (Transparent Display): Japan Display Inc.


ContactInquiries /

At xDiversity, we welcome you to contact us
if you would like to participate in our research as
a tester or user
or if you are a business
with data that can be used.
Please contact us via the form below.

Research collaborator recruitment form

Other Inquiries