International Display Workshops General Incorporated Association

4:10 PM - 6:10 PM

[3Dp1-7L] Development of Spoken Language Identification System Using Directional Volumetric Display

*Mitsuru Baba1, Naoto Hoshikawa2, Hirotaka Nakayama3, Tomoyoshi Ito1, Atsushi Shiraki1 (1. Chiba University(Japan), 2. National Institute of Technology, Oyama College(Japan), 3. National Astronomical Observatory of Japan(Japan))

Volumetric displays, Digital signage, Machine learning

https://doi.org/10.36463/idw.2020.0519

In this study, we develop the spoken language identification system by a directional volumetric display. We proposed the identification model using CNN for four languages, English, French, German, and Spanish. The identification accuracy was 93.4%, and became possible to display an image in the observer's direction.