Voice-Controlled Audio Adjustment for Accurate Speech Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Automatic speech recognition (ASR) in voice-controlled devices is hindered by background noise, making it difficult to accurately identify voice commands within audio signals.
Innovation Solution
The device outputs audio and captures sound to identify a predefined phrase, then alters the audio output to reduce noise by attenuating, pausing, or switching to mono, based on user distance, direction, and audio type, increasing the signal-to-noise ratio for improved speech recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If audio output is maintained at full volume, then entertainment function is preserved, but speech recognition accuracy deteriorates due to background noise
Solution Approach 1:
The system dynamically adjusts the audio output based on detected speech events. When speech is detected, the audio output is temporarily modified (attenuated, paused, or switched to mono) to reduce background noise during critical recognition moments. This dynamic adjustment resolves the contradiction by making the audio output adaptive rather than static, allowing full volume during non-speech periods and reduced noise during speech periods.
Solution Approach 2:
The system changes audio output parameters (volume level, stereo/mono configuration) based on the detected speech context. By modifying these parameters temporarily when speech is detected, the system reduces the harmful background noise effect while maintaining entertainment functionality during non-speech periods. This parameter adjustment directly addresses the contradiction between reliability and harmful factors.
2Measurement precision
If audio is attenuated or paused during speech detection, then signal-to-noise ratio improves, but user entertainment experience deteriorates
Solution Approach 1:
The system applies periodic modulation to the audio output, temporarily attenuating or pausing audio during detected speech periods and restoring full output between speech events. This periodic action ensures that speech recognition occurs during high signal-to-noise ratio periods while entertainment continues during non-speech periods, resolving the contradiction between measurement precision and ease of operation.
Solution Approach 2:
The audio output transitions from static full-volume playback to dynamic adjustment based on speech detection. The system continuously monitors for speech and adapts the audio output in real-time, creating a dynamic experience that prioritizes recognition accuracy during speech moments while maintaining entertainment quality during non-speech moments.
3Reliability
If stereo audio is switched to mono during speech recognition, then noise reduction is achieved, but audio quality deterioration occurs
Solution Approach 1:
The system periodically switches between stereo and mono audio output based on speech detection. During detected speech periods, audio is switched to mono to reduce noise and improve recognition reliability. Between speech events, full stereo quality is restored to maintain audio quality for entertainment. This periodic switching resolves the contradiction between reliability and manufacturing precision.
Solution Approach 2:
The audio output configuration dynamically switches between stereo and mono modes based on the detected speech context. This dynamic mode switching allows the system to optimize for recognition reliability during speech moments while preserving audio quality during non-speech entertainment periods.
Data Source
AI summary
Techniques for altering audio being output by a voice-controlled device, or another device, to enable more accurate automatic speech recognition (ASR) by the voice-controlled device. For instance, a voice-controlled device may output audio within an environment using a speaker of the device. While outputting the audio, a microphone of the device may capture sound within the environment and may generate an audio signal based on the captured sound. The device may then analyze the audio signal to identify speech of a user within the signal, with the speech indicating that the user is going to provide a subsequent command to the device. Thereafter, the device may alter the output of the audio (e.g., attenuate the audio, pause the audio, switch from stereo to mono, etc.) to facilitate speech recognition of the user's subsequent command.