Electronic device for identifying driver on basis of multimodal data including various types of data related to vehicle drive

The electronic device addresses the limitations of existing systems by using multimodal data and Cross-Attention algorithms to enhance driver identification accuracy in complex scenarios.

WO2026146701A1PCT designated stage Publication Date: 2026-07-09IND ACADEMIC COOP FOUND HALLYM UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
IND ACADEMIC COOP FOUND HALLYM UNIV
Filing Date
2025-01-22
Publication Date
2026-07-09

AI Technical Summary

Technical Problem

Existing driver analysis systems fail to adequately address complex driving scenarios due to reliance on single data, failing to effectively reflect the need to identify drivers by considering all types of data, and existing systems, and existing systems have limitations in effectively identifying drivers by considering all types of data capable of verifying driving patterns.

Method used

An electronic device that utilizes a processor to input multimodal data, including image data, sensing data, and CAN-bus data into multiple driver identification models, employing Cross-Attention algorithms to generate prediction values, and ensemble these values to accurately identify drivers.

Benefits of technology

The system effectively identifies drivers by considering multiple data types, enhancing accuracy and reflecting complex driving scenarios through multimodal data analysis.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure KR2025001250_09072026_PF_FP_ABST
    Figure KR2025001250_09072026_PF_FP_ABST
Patent Text Reader

Abstract

Disclosed is an electronic device for identifying a driver on the basis of multimodal data. The electronic device according to the present disclosure comprises: a memory for storing at least one driver identification model; and a processor connected to the memory and for inputting, to the at least one driver identification model, multimodal data acquired during a driving process of a vehicle and acquiring prediction data for the driver of the vehicle on the basis of the output of the driver identification model, wherein the multimodal data includes a plurality of pieces of image data obtained by imaging the outside of a traveling vehicle, sensing data related to the movement of the driving vehicle, and controller area network (CAN)-bus data related to the state of the traveling vehicle.
Need to check novelty before this filing date? Find Prior Art