Audio Interval Detection Using Phoneme Segmentation to Filter Noise
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio interval detection methods often erroneously detect noise as speech utterance intervals, leading to unnecessary resource consumption and missed speech data, due to low detection accuracy.
Innovation Solution
An audio interval detection apparatus that identifies a specified audio interval representing a continuous phoneme state over a certain time and eliminates it to detect the utterance audio interval, using a neural network to differentiate between speech and noise signals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional utterance audio interval detection is used, then detection coverage is achieved, but detection accuracy deteriorates due to erroneous detection of noise as speech
Solution Approach 1:
The audio signal is segmented into multiple frames, and phoneme detection is performed on each frame independently. The phoneme detection unit divides the utterance audio interval into multiple frames and detects phonemes in each frame, allowing for precise identification of speech segments while excluding noise intervals that don't contain valid phoneme patterns
Solution Approach 2:
A phoneme detection unit is introduced as an intermediary between the utterance audio interval detection and the final speech identification. This unit detects phonemes in each frame and determines whether each frame contains a phoneme, providing an intermediate verification layer that filters out noise intervals before final speech detection
2Measurement precision
If utterance audio interval detection is performed on entire audio data, then comprehensive speech detection is achieved, but resource consumption increases unnecessarily
Solution Approach 1:
The phoneme detection unit extracts only the essential phoneme information from each frame, determining whether a phoneme is present without performing full speech analysis on entire audio segments. This extraction approach allows the system to identify potential speech intervals efficiently before applying more resource-intensive voice recognition only where needed
Solution Approach 2:
The system performs partial phoneme detection on individual frames rather than complete speech analysis on entire audio segments. By detecting phonemes frame-by-frame and identifying only intervals containing phonemes as potential speech, the system achieves comprehensive speech detection coverage while minimizing resource consumption by avoiding unnecessary analysis of noise intervals
3Measurement precision
If detection threshold is lowered to catch more speech, then detection sensitivity improves, but false detection of noise increases
Solution Approach 1:
The system uses feedback from phoneme detection results to adjust speech interval identification. The phoneme detection unit provides feedback on whether each frame contains a phoneme, and this information is used to determine valid speech intervals. This feedback mechanism allows the system to maintain high sensitivity while filtering out noise through iterative verification
Data Source
AI summary
An audio interval detection apparatus has a processor and a storage storing instructions that, when executed by the processor, control the processor to: detect, from a target audio signal, a specified audio interval including a specified audio signal representing a state of a phoneme of a same consonant produced continuously over a period longer than a specified time, and, by eliminating, from the target audio signal at least the detected specified audio interval, detect from the target audio signal an utterance audio interval that includes a speech utterance signal representing a speech utterance uttered by a speaker.