Recently, Li Zhengguo, a senior researcher at the Suzhou Research Institute of the National University of Singapore, published scientific research results in the international top academic journal Nature Communications. Li Zhengguo's team successfully developed a sign language recognition and communication system based on triboelectric gloves, which not only realizes the universal recognition of multiple sign language words with the assistance of artificial intelligence, but also analyzes and translates sentences, especially new sentences that have not been trained by artificial intelligence, to meet the daily communication needs of people with language and hearing impairment who use sentences with high frequency, and improve the practicality of the sign language recognition system.
The system developed by Li Zhengguo's team includes triboelectric sensing gloves, deep learning modules, and a user interface. As a demonstration, the system achieves independent recognition of 50 words and 20 sentences, with an accuracy of up to 91.3% and 95%, respectively, within a non-segmentation framework. In addition, in order to overcome the limitation of not being able to recognize new sentences without training, the segmentation method divides the sentence signals in the dataset into word fragments, and the deep learning model is responsible for memorizing all the word units that have appeared. At the same time, on the virtual reality interface, people with language and hearing impairment can directly use sign language to interact with non-sign language users, and non-sign language users can directly type and participate in communication. "In addition to improving the social participation of people with language and hearing impairments, this technology can also be applied to assistive devices for the disabled, medical and health care, etc., to help people-oriented Internet of Things construction." Li Zhengguo said.
【Source: Suzhou Daily】
Copyright belongs to the original author, a tribute to the original
![The research results of Xifeng Liquor have passed the expert appraisal[fig]](data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=)