Speech Evaluation Apparatus Using Cutoff Frequency Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for improving speech generation through lowpass filtering and modified auditory feedback do not clearly define the type of feedback required for improvement and lack a systematic approach for utterance training.
Innovation Solution
An evaluation and training device that analyzes speech signals to determine formant frequencies and applies lowpass filters with specific cutoff frequencies, feeding back converted signals to subjects to calculate compensatory responses and evaluate perception characteristics and train utterance effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If lowpass filter with cutoff frequency of 2 kHz or lower is applied to feedback speech signal, then speech intelligibility is improved, but voice identity recognition deteriorates
Solution Approach 1:
The patent applies parameter changes by using different cutoff frequencies (2 kHz and 8 kHz) for lowpass filtering feedback speech signals. This allows the system to control the balance between speech intelligibility improvement and voice identity information preservation by adjusting the cutoff frequency parameter.
Solution Approach 2:
The patent segments the evaluation process into multiple conditions: one with lowpass filtering (2 kHz cutoff) to evaluate speech intelligibility improvement, and another without lowpass filtering or with higher cutoff (8 kHz) to evaluate voice identity recognition. This segmentation allows independent assessment of each aspect.
2Adaptability or versatility
If formant frequency conversion is applied to feedback speech signal, then compensatory response is induced, but the type of feedback required for improvement is not clearly defined
Solution Approach 1:
The patent uses formant frequency conversion as a parameter change technique to induce compensatory responses in subjects. By converting formant frequencies in the feedback speech signal, the system creates controlled auditory feedback conditions that elicit measurable compensatory responses while maintaining a relatively simple system architecture.
3Measurement precision
If multiple cutoff frequencies are used for evaluation, then perception characteristics can be comprehensively evaluated, but the evaluation process becomes more complex
Solution Approach 1:
The patent segments the evaluation into distinct conditions using different cutoff frequencies (2 kHz and 8 kHz). Each condition evaluates specific aspects of perception characteristics independently, allowing comprehensive evaluation while maintaining clear separation between test conditions to manage complexity.
Solution Approach 2:
The patent implements feedback by providing subjects with speech signals under different filtering conditions and measuring their compensatory responses. The feedback mechanism compares performance across conditions to evaluate perception characteristics systematically.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables the evaluation of perception characteristics and training of utterances by establishing a clear relationship between cutoff frequencies and compensatory responses, improving speech generation and recognition of voice identity.
Implementation Method 1
a conversion unit that applies a lowpass filter with a cutoff frequency being a first predetermined value or a second predetermined value greater than the first predetermined value with or without change of feedback formant frequencies
Data Source
AI summary
An evaluation device applies a lowpass filter with a cutoff frequency being a first predetermined value or a second predetermined value greater than the first predetermined value with or without change of feedback formant frequencies which are formant frequencies of a picked-up speech signal, converts the picked-up speech signal, feeds back the converted speech signal to a subject, and includes an evaluation unit that calculates a compensatory response vector by using pickup formant frequencies which are formant frequencies of a speech signal acquired by picking up an utterance made by the subject while feeding back a speech signal that has been converted with change of the feedback formant frequencies to the subject, and pickup formant frequencies which are formant frequencies of a speech signal acquired by picking up an utterance made by the subject while feeding back a speech signal that has been converted without change of the feedback formant frequencies to the subject, and determines an evaluation based on a compensatory response vector for each cutoff frequency.