Naked eye 3D vision training method and system based on AI neural network

Through the naked-eye 3D vision training method based on AI neural network, the features are extracted using MTCNN and CNN networks, the fusion cascade cost body is constructed, and the training model is optimized, which solves the problems of insufficient data and high computing resources in 3D vision training, and achieves a more efficient training effect.

CN120267502APending Publication Date: 2025-07-08ZHEJIANG UNIV OF TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202510337927.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-03-21
Publication Date
2025-07-08

AI Technical Summary

Technical Problem

The prior art lacks high-quality training data in 3D vision training, and traditional methods ignore the laws of physical motion, which leads to high difficulty in model training and high computing resource consumption.

Method used

The naked-eye 3D vision training method based on AI neural network is adopted. By obtaining the identity and naked-eye information of the person to be trained, the eye images are collected using MTCNN and naked-eye 3D tracking technology, and the features are extracted by combining the CNN network and attention mechanism, a fusion cascade cost body is constructed for parallax prediction, and the training model is optimized.

Benefits of technology

It improves binocular regulation function and brain network efficiency, reduces computing resource consumption, improves the accuracy and efficiency of training results, reduces the variability of dynamic brain networks, and enhances the coupling of neural activity and hemoglobin.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120267502A_ABST
    Figure CN120267502A_ABST
Patent Text Reader

Abstract

The invention discloses a naked eye 3D visual training method and system based on an AI neural network, and the method comprises the steps: S1, obtaining the identity information and naked eye information of a to-be-trained person, and enabling the naked eye information to comprise refraction data and the strabismus degree of two eyes; s2, performing difference matching on the naked eye information to obtain preprocessed data; s3, constructing an initial training model based on an artificial intelligence neural network, and inputting the preprocessed data into the initial training model to generate a personal training scheme; and S4, training the naked eye 3D vision based on the generated training scheme, and outputting a training result.
Need to check novelty before this filing date? Find Prior Art