Incremental Speaker Enrollment via Temporary Voice Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing speaker systems in commercial robots face determination errors and incorrect command voice execution due to similarity threshold issues in incremental enrollment, leading to deviated voiceprint features and reduced identification precision over time.
Innovation Solution
An electronic apparatus with a microphone, storage device, and processor that temporarily stores command voices with similarity scores above a threshold and applies screening rules to select incremental enrolled voices only after a predetermined number exceeds a value, enhancing precision by excluding incorrect voices and maintaining accurate voiceprint features.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the system directly stores command voices as enrolled voices when similarity exceeds a threshold, then the enrollment process is simple and fast, but determination errors occur and voiceprint features deviate over time
Solution Approach 1:
The patent introduces a temporary storage library as an intermediate step before final enrollment. Command voices that exceed the similarity threshold are first stored temporarily, then subjected to secondary verification through comparison with already enrolled voices. Only after passing this preliminary filtering step are they permanently enrolled, thus preventing premature incorrect enrollments while maintaining efficient processing.
Solution Approach 2:
The temporary storage library acts as an intermediary between the threshold-based initial filtering and the final enrollment database. This intermediate layer allows for additional verification without blocking the main enrollment flow, serving as a buffer that reconciles the conflict between speed and accuracy by enabling batch verification of temporarily stored voices.
2Measurement precision
If the system uses a high similarity threshold for enrollment, then identification precision is maintained, but incorrect command voices are still executed and stored due to determination errors
Solution Approach 1:
The patent implements a feedback mechanism where temporarily stored voices are continuously compared against the enrolled voice database. The system uses the accumulated similarity scores from multiple comparisons to dynamically adjust enrollment decisions. This feedback loop allows the system to recognize patterns and reduce determination errors, as incorrect voices that consistently fail to match enrolled voices are filtered out even if they initially exceeded the threshold.
Solution Approach 2:
The system applies excessive verification by comparing temporarily stored voices multiple times against the enrolled database before final enrollment. Rather than relying on a single threshold check, the system performs repeated comparisons and only enrolls voices that consistently demonstrate high similarity across multiple evaluations, thus reducing determination errors through redundant verification.
3Reliability
If the system performs multiple verification steps before enrollment, then enrollment accuracy improves, but the complexity of the enrollment process increases
Solution Approach 1:
The patent segments the enrollment process into distinct stages: initial threshold-based filtering, temporary storage, secondary verification against enrolled voices, and final enrollment. This segmentation allows each stage to have a specific, simple function rather than one complex monolithic process. The temporary storage library serves as a clear boundary between the simple initial filtering and the more rigorous verification stage.
Solution Approach 2:
The system dynamically adjusts the enrollment process based on the characteristics of the command voice. Voices that clearly exceed the threshold are quickly enrolled through the streamlined path, while those near the threshold undergo more extensive verification. This dynamic approach allows the system to adapt its complexity to the needs of each specific enrollment case, avoiding unnecessary complexity for clear-cut cases while applying rigorous verification only when needed.
Data Source
AI summary
The present disclosure provides an electronic apparatus having an incremental enrollment unit and a method thereof. The electronic apparatus at least includes a microphone, a storage device, and a processor. The storage device stores a first screening rule, an enrollment database, and a first temporary storage library. The processor receives a command voice transmitted by the microphone, and compare the command voice with enrolled voices in the enrollment database. If determining that a similarity is larger than a threshold value, the processor stores the command voice as a first temporarily stored voice in the first temporary storage library. When a quantity of the first temporarily stored voices in the first temporary storage library is larger than a first predetermined value, the processor screens out a part of the first temporarily stored voices according to the first screening rule, so as to perform incremental enrollment.