Method and system for collecting decision data by multi-modal large model driven intelligent agent

Through the multimodal large model-driven intelligent body method, intelligent trolleys can efficiently integrate multimodal information and optimize decision data in complex environments, solving the challenges of existing systems in decision efficiency and accuracy, and achieving more accurate and efficient task execution.

CN119808006AActive Publication Date: 2025-04-11NANJING ARTIFICIAL INTELLIGENCE CHIPS RES INST OF AUTOMATION CHINESE ACAD OF SCI +1
3 Cites 28 Cited by

Patent Information

Application Number
CN202510287600.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-03-12
Publication Date
2025-04-11
Estimated Expiration
2045-03-12

AI Technical Summary

Technical Problem

The existing smart car decision-making system has efficiency and accuracy challenges in converging multimodal information and optimizing decision data, especially in complex environments and multi-round interactive scenarios.

Method used

Adaptive multimodal large model-driven agent method is adopted to obtain sensor data, calculate information entropy and normalized mass scores, and adaptive multimodal information fusion is performed to generate fusion feature vectors. Then, time-sensitive multimodal alignment processing is applied, combining timestamp embedding and timing self-attention mechanisms to generate timing alignment features. Based on these features, a Bayesian neural network is used to generate a decision distribution, assess uncertainty and generate a final decision and indicator set.

Benefits of technology

It realizes the efficient decision-making process of smart cars in complex environments. Through multiple rounds of interaction, the decision-making accuracy and execution capabilities are improved, ensuring the accuracy of task completion and the system's response capabilities.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention discloses a method and a system for collecting decision data by an intelligent agent driven by a multi-modal large model. The method comprises the following steps: collecting original data of a predetermined sensor, calculating information entropy and a normalized mass score, and generating a fusion feature vector through self-adaptive multi-modal information fusion; time sequence sensitive multi-mode alignment processing is carried out on the image, and time sequence alignment features are obtained by combining timestamp embedding and a self-attention mechanism; based on this, decision distribution is generated by using a Bayesian neural network, uncertainty is evaluated, and a final decision and index set is generated; evaluating a task score in combination with task description, and making an execution plan according to a resource state; initializing the model, generating synthetic data, optimizing the model and storing experience. According to the invention, the efficient decision-making process of the intelligent trolley is realized, and the execution capability and task completion precision of the intelligent trolley are improved.
Need to check novelty before this filing date? Find Prior Art