Title Microwave Lip Reading of Chinese Mandarin Based on Programmable Metasurface
Authors Jiang, Siyuan
Ruan, Hengxin
Wang, Zhuo
Zhang, Hongrui
Zhao, Hanting
Li, Lianlin
Affiliation Peking Univ, Dept Elect, Beijing, Peoples R China
PengCheng Natl Labrotory, Shenzhen, Peoples R China
Issue Date 2021
Publisher 2021 IEEE MTT-S INTERNATIONAL MICROWAVE WORKSHOP SERIES ON ADVANCED MATERIALS AND PROCESSES FOR RF AND THZ APPLICATIONS (IMWS-AMP)
Abstract Lip reading is an important way of communication in some specific environment. The traditional lip reading system, which is based on processing and recognizing optical video is becoming not that applicable in areas where the light is obstructed. In order to overcome the shortcomings of optical and video-based methods, microwave lip reading was proposed. The current microwave lip reading system is capable of processing lip reading tasks in the English context. However, this method is not able to distinguish the lip language at the syllable level. In this work, we choose Chinese character as the object of syllable classification task, because each Chinese character consists of only one syllable. We utilize the programmable metasurface to focus the EM wave emitted by the transmitter to the speaker's lip and get the microwave data which contains information of the lip's movement. Then, we use deep learning method to classify the different Chinese character with very high accuracy. At the same time, we prove that the microwave lip reading can detect the movement inside one's mouth cavity. As a result, the microwave lip reading system is able to classify the syllables more precisely than optical method.
URI http://hdl.handle.net/20.500.11897/647073
ISBN 978-1-6654-2819-4
DOI 10.1109/IMWS-AMP53428.2021.9643862
Indexed EI
CPCI-S(ISTP)
Appears in Collections: 电子学院

Files in This Work
There are no files associated with this item.

Web of Science®


0

Checked on Last Week

Scopus®



Checked on Current Time

百度学术™


0

Checked on Current Time

Google Scholar™





License: See PKU IR operational policies.