Intelligent glasses system and method for multi-modal bio-signal intent recognition

By using a multimodal biosignal recognition system that combines eye movement, electromyography, and environmental perception, the smart glasses achieve covert, accurate, and low-load user intent recognition, solving the problems of interaction privacy, low recognition rate, and misoperation in existing technologies, and supporting long-term use.

CN122111222APending Publication Date: 2026-05-29JIANGXI XINSHIJIA OPTOELECTRONICS TECHNOLOGY CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
JIANGXI XINSHIJIA OPTOELECTRONICS TECHNOLOGY CO LTD
Filing Date
2026-01-30
Publication Date
2026-05-29

AI Technical Summary

Technical Problem

Existing contactless interaction technologies for smart glasses suffer from issues such as privacy leaks in voice interaction, low recognition rates, significant impact of ambient light on gesture interaction, and misoperations in eye-tracking, making it difficult to achieve covert, accurate, and interference-resistant user intent recognition.

Method used

A multimodal biosignal recognition system is adopted, including an eye-tracking acquisition unit, a bioelectrical signal acquisition unit, and an environment and situation sensing unit. It combines a signal preprocessing engine and a multimodal intent prediction model, and utilizes a flexible dry electrode array, an inward infrared corneal reflectometer, a 6-axis IMU, a VAD sensor, and a SLAM scene camera. Through a cross-attention mechanism, eye-tracking and electromyographic features are fused to achieve intent recognition.

Benefits of technology

Users can complete the operation with a slight biting motion, reducing barriers to use in social situations, with a low rate of false triggering, supporting long-term use, and improving the concealment and accuracy of the interaction.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122111222A_ABST
    Figure CN122111222A_ABST
Patent Text Reader

Abstract

The application provides a kind of multi-modal biological signal's intelligent glasses intention recognition system and method, it is related to intelligent glasses technical field, including multidimensional perception module, signal preprocessing engine, multi-modal intention prediction model and feedback execution layer;Multidimensional perception module contains eye movement acquisition unit, bioelectricity acquisition unit and environment and situation sensing unit;Signal preprocessing engine is connected to multidimensional perception module, for noise filtering and signal logic gate control;Multi-modal intention prediction model is based on cross attention mechanism fusion eye movement and myoelectricity characteristics;Feedback execution layer drives AR display interface interaction according to prediction result;Wherein, bioelectricity acquisition unit uses flexible dry electrode array, is distributed in temporal muscle and after ear muscle group area, environment and situation sensing unit integrates 6-axis IMU, VAD sensor and SLAM scene camera.The application can keep interaction concealment, can accurately identify user real operation intention, and has strong anti-interference ability.
Need to check novelty before this filing date? Find Prior Art