Nonphonetic Audio Command Detection in Voice Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Voice-activated systems face difficulties in recognizing voice commands in noisy environments, particularly when background noise from conversations or other sources masks intended phonetic commands, leading to degraded performance.
Innovation Solution
A sound-activated control system that includes a receiver and a command discriminator to detect temporally and/or spectrally compact nonphonetic audio commands within digital audio waveforms, which can control voice-activated systems by synthesizing phonetic commands when necessary.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If voice-activated systems use phonetic commands for control, then the system can respond to natural speech, but the system performance degrades in noisy environments with background conversations
Solution Approach 1:
The patent extracts and removes background noise from the audio signal using spectral subtraction and noise modeling techniques. The system identifies and separates the desired phonetic command from background conversations by estimating noise characteristics and subtracting them from the received signal, thereby improving command recognition accuracy in noisy environments
Solution Approach 2:
The patent introduces an intermediary noise estimation and spectral subtraction process between the noisy audio input and the voice command recognition system. This intermediary layer processes the audio signal to enhance the phonetic command by removing background noise components, allowing the recognition system to operate more effectively
2Measurement precision
If the system filters out background noise to improve command detection, then recognition accuracy improves, but the system complexity increases
Solution Approach 1:
The patent changes the parameters of the audio signal by transforming it from the time domain to the frequency domain using Fourier transform. This parameter transformation enables more effective noise filtering and command extraction by operating in the spectral domain where noise and signal components can be separated based on their frequency characteristics
Data Source
AI summary
A sound-activated control system includes an audio receiver and a command discriminator. The receiver is configured to receive an audio waveform and to produce a digital audio waveform therefrom. The command discriminator is configured to detect a temporally and/or spectrally compact nonphonetic audio command within the digital audio waveform and to control a voice-activated system an action in response to the nonphonetic command.