Voice Recognition System User Characteristic Adaptation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current voice recognition systems in electronic devices face challenges in accurately recognizing user voices due to variations in characteristics such as sex and age, which affect tone, pitch, and other acoustic features, leading to errors in command interpretation.
Innovation Solution
A voice recognition method and apparatus that determine user characteristics by analyzing input voice features like frequency, tempo, pitch, and perturbations, and compare them to a voice-model database to generate responsive voices and control devices accordingly, taking into account sex and age-specific attributes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a general voice recognition system is used without considering user characteristics, then the device complexity is reduced, but the voice recognition accuracy deteriorates due to variations in tone and pitch across different users
Solution Approach 1:
The system performs preliminary voiceprint registration during a setup phase, storing characteristic data (pitch, tone, frequency) for each user before actual voice recognition operations. This preliminary action creates user-specific reference models that enable accurate recognition without adding complexity to the real-time recognition process.
Solution Approach 2:
The system changes recognition parameters dynamically based on detected user characteristics. By analyzing pitch, tone, and frequency parameters of the input voice and comparing them against stored voiceprints, the system adapts its recognition thresholds and parameters to match the specific user, thereby improving accuracy without requiring a completely different system architecture.
2Reliability
If voice recognition takes into account user characteristics such as sex and age, then the voice recognition accuracy is improved, but the device complexity increases due to additional processing requirements
Solution Approach 1:
The voice recognition process is segmented into distinct phases: voiceprint registration phase where user characteristics are captured and stored, and voice recognition phase where input voices are compared against stored prints. This segmentation allows the system to handle complex characteristic analysis only during the initial registration, keeping ongoing recognition operations simpler and more reliable.
Solution Approach 2:
The system creates simplified copies of user voice characteristics in the form of voiceprint data structures that store essential parameters (pitch, tone, frequency ranges). These copied representations enable reliable comparison and recognition without requiring the system to re-analyze complex acoustic features during each recognition operation.
3Measurement precision
If the system analyzes multiple voice features such as frequency, tempo, pitch, and perturbations, then the user characteristics determination accuracy is improved, but the measurement precision requirements increase
Solution Approach 1:
The system analyzes multiple voice features (frequency, tempo, pitch, perturbations) but implements them in a progressive manner. During voiceprint registration, comprehensive analysis of all features is performed to capture complete user characteristics. During subsequent recognition operations, the system uses these pre-analyzed features stored in the voiceprint, reducing the measurement burden while maintaining high determination accuracy.
Data Source
AI summary
Disclosed are an electronic apparatus and a voice recognition method for the same. The voice recognition method for the electronic apparatus includes: receiving an input voice of a user; determining characteristics of the user; and recognizing the input voice based on the determined characteristics of the user.