Automated ADE Score Extraction from Vehicle Speech Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for evaluating vehicle performance in automatic driving are not accurate and have low collection efficiency, leading to larger errors in performance assessment.
Innovation Solution
A vehicle performance evaluation method that acquires labeled auto driving experience (ADE) scores and data, calculates correlations between ADE items and vehicle states, and generates evaluation results without passenger participation, using modules for scoring, data recording, correlation calculation, and evaluation generation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If passenger scoring is used to evaluate ADE, then evaluation data can be collected, but collection efficiency is low and accuracy is poor
Solution Approach 1:
The patent replaces the mechanical manual scoring system with an automated speech recognition and analysis system. The speech processing module automatically transcribes and analyzes passenger speech to extract ADE scores, eliminating the need for manual scoring while improving both efficiency and consistency of data collection
Solution Approach 2:
The system enables self-service evaluation by automatically processing passenger feedback without requiring manual intervention. The speech-to-text conversion and automated analysis allow the system to evaluate ADE scores independently, improving collection efficiency while maintaining evaluation accuracy
2Reliability
If manual scoring method is used, then ADE scores can be obtained, but time consumption is high and costs increase
Solution Approach 1:
The system performs preliminary action by pre-processing and storing speech data during the test drive, then automatically analyzing it afterward. This allows the evaluation process to be decoupled from the actual driving test, significantly reducing the time required for score generation while maintaining reliability through consistent automated analysis
Solution Approach 2:
Manual scoring operations are replaced with automated speech recognition and natural language processing systems. This substitution eliminates human time consumption and subjectivity, providing reliable automated evaluation that operates continuously without fatigue or inconsistency
3Loss of information
If passenger participation is required for scoring, then ADE data can be collected, but operational complexity increases
Solution Approach 1:
The system extracts ADE evaluation information directly from passenger speech without requiring structured input formats or active participation. By taking out only the relevant speech data and automatically analyzing it, the system maintains data completeness while greatly simplifying the operational process for passengers
Solution Approach 2:
The automated speech processing system performs the evaluation function independently without requiring passenger training or active cooperation. Passengers simply speak naturally while the system self-services the entire evaluation process, from data collection to score generation
Data Source
AI summary
A vehicle performance evaluation method, device and terminal are provided. The method includes: acquiring a labeled ADE score of an ADE item within a time period in which the ADE item occurs, and recording labeled data of an ADE index for indicating a vehicle state; acquiring a correlation between the ADE item and the vehicle state, according to the labeled ADE score and the labeled data of the ADE index; acquiring data of a target ADE index within a preset time period, and acquiring a target ADE score according to the data of the target ADE index and a correlation between the ADE item and the vehicle state; and acquiring a vehicle performance evaluation result according to the target ADE score. The passenger participation is not necessary, costs can be reduced and a vehicle evaluation efficiency can be improved.