An electroencephalogram signal recognition and adjustment method and system based on artificial intelligence
By using an AI-based signal trimming and quantization model, the problems of high power consumption and insufficient accuracy in traditional EEG signal recognition methods have been solved, achieving high-precision and low-power EEG signal recognition, which can be applied in multiple fields.
CN122153359APending Publication Date: 2026-06-05UNIV OF MACAU
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- UNIV OF MACAU
- Filing Date
- 2026-01-23
- Publication Date
- 2026-06-05
AI Technical Summary
Technical Problem
Traditional EEG signal recognition methods suffer from high power consumption and insufficient recognition accuracy when using high-bit quantization, while 4-bit quantization cannot meet the high accuracy requirements.
Method used
An AI-based signal cropping and signal recognition model is used to crop and quantize EEG signals through a training network, discarding abnormal or high-amplitude sample points. Quantization is combined with an analog front-end to reduce recognition power consumption and improve accuracy.
Benefits of technology
It effectively improves the accuracy and speed of EEG signal recognition, reduces recognition power consumption, and is applicable to fields such as medical health, human-computer interaction, consumer electronics, and neuroscience research.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN122153359A_ABST
Abstract
The application discloses an electroencephalogram signal recognition and adjustment method and system based on artificial intelligence. The electroencephalogram signal recognition and adjustment method comprises the following steps: collecting an electroencephalogram signal to be processed; training a preset group network to obtain a task processing model, wherein the task processing model comprises a signal cutting model and a signal recognition model; cutting the electroencephalogram signal to be processed through the signal cutting model to obtain a first electroencephalogram signal; quantifying the first electroencephalogram signal through an analog front end to obtain a second electroencephalogram signal; and identifying the second electroencephalogram signal through the signal recognition model to obtain a signal recognition result. The application can reduce the recognition power consumption of the signal recognition model, effectively improve the recognition accuracy and speed of the electroencephalogram signal, and can be widely applied to the technical field of image processing.
Need to check novelty before this filing date? Find Prior Art