User Time Markers for Noise-Robust Speech Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automated speech recognition (ASR) systems struggle with robustness in the presence of background noise and competing speech, leading to inaccurate transcriptions, particularly in low signal-to-noise ratio conditions.
Innovation Solution
The use of a sequence of time markers input by the user in cadence with their speech, correlated with the audio signal to enhance audio features, separating target speech from background noise and competing speech using a speech recognition model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional ASR systems process audio signals directly, then the system complexity remains low, but the recognition accuracy deteriorates in noisy environments
Solution Approach 1:
The system performs preliminary actions by having users input time markers before speech recognition occurs. These time markers are used to pre-segment the audio signal into speech and non-speech portions, preparing the data in advance for more accurate recognition processing.
Solution Approach 2:
The audio signal is segmented into speech segments and non-speech segments based on the time markers provided by the user. This segmentation allows the ASR system to focus processing resources on relevant speech portions while excluding background noise and competing speech.
2Reliability
If ASR systems operate without time markers, then the ease of operation is high, but the reliability of speech recognition deteriorates under background noise
Solution Approach 1:
The system leverages the user's own temporal awareness of their speech to create the time markers. Users naturally know when they speak and can easily indicate these moments, making the process intuitive and self-service oriented rather than requiring complex external tools.
Solution Approach 2:
The time markers provided by the user serve as feedback to the ASR system about when speech occurs. This feedback mechanism allows the system to adjust its processing to focus on relevant time periods, significantly improving reliability in noisy environments.
3Measurement precision
If the system processes all audio segments including background noise, then the processing speed is high, but the transcription accuracy deteriorates
Solution Approach 1:
The system extracts only the relevant speech segments from the complete audio signal using time markers. By taking out and processing only the portions containing actual speech while excluding background noise and competing speech, the system achieves both high accuracy and efficient processing.
Data Source
AI summary
A method for separating target speech from background noise contained in an input audio signal includes receiving the input audio signal captured by a user device, wherein the input audio signal corresponds to target speech of multiple words spoken by a target user and containing background noise in the presence of the user device while the target user spoke the multiple words in the target speech. The method also includes receiving a sequence of time markers input by the target user in cadence with the target user speaking the multiple words in the target speech, and correlating the sequence of time markers with the input audio signal to generate enhanced audio features that separate the target speech from the background noise in the input audio signal. The method also includes processing, using a speech recognition model, the enhanced audio features to generate a transcription of the target speech.