Playback-Triggered Command Keyword Windows for False-Positive Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional wake-word engines in network microphone devices are prone to false positives due to false wake words and phonetically similar words, leading to resource consumption and privacy concerns from continuous listening and data transmission.
Innovation Solution
Implementing a command keyword engine that detects command keywords during a predetermined input detection window triggered by media item changes, requiring both keyword detection and specific playback conditions before executing commands, and utilizing local natural language processing to reduce false positives and enhance privacy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If wake-word engines continuously listen for commands, then command responsiveness is improved, but false positives increase and privacy is compromised
Solution Approach 1:
The system performs preliminary actions by detecting media item changes (track transitions, playback state changes) before opening the input detection window. This preliminary detection allows the system to be ready for commands at appropriate moments without continuously listening, thereby reducing false positives while maintaining responsiveness.
Solution Approach 2:
Instead of continuous listening, the system uses periodic action by opening input detection windows only during specific periods when media items are playing. The wake-word engine activates periodically during playback events rather than continuously, which reduces false positives from background noise while maintaining command responsiveness during relevant time windows.
2Adaptability or versatility
If wake-word engines continuously transmit audio data to cloud servers, then processing capability is improved, but data privacy is compromised and resource consumption increases
Solution Approach 1:
The system extracts and processes audio data locally on the playback device before transmitting to cloud servers. The wake-word engine and natural language processing occur locally, extracting only necessary information (such as command keywords) for cloud processing, thereby reducing the volume of data transmitted and minimizing privacy risks while maintaining processing capability.
Solution Approach 2:
The system introduces local processing components (wake-word engine, natural language processing module) as intermediaries between the microphone and cloud server. These intermediaries filter and preprocess audio data locally, acting as a mediator that reduces the information transmitted to the cloud while maintaining the ability to process commands, thus balancing processing capability with privacy protection.
3Measurement precision
If input detection windows are opened continuously, then command detection accuracy is improved, but resource consumption increases
Solution Approach 1:
The system applies dynamics by making the input detection window adaptive rather than static. The window opens dynamically in response to media item changes (track transitions, playback state changes) and closes when media playback stops. This dynamic approach maintains command detection accuracy during relevant periods while significantly reducing resource consumption during inactive periods.
Solution Approach 2:
The system changes the parameter of detection window duration and frequency based on media playback state. During active playback, the detection window is open with appropriate duration to capture commands. When playback stops or during non-relevant periods, the detection window closes. This parameter change strategy maintains detection accuracy when needed while reducing overall resource consumption.
Data Source
AI summary
A device, such as Network Microphone Device or a playback device, receives an indication of a track change associated with a playback queue output by a media playback system. In response, an input detection window is opened for a given time period. During the given time period the device is arranged to receive an input sound data stream representing sound detected by a microphone. The input sound data stream is analyzed for a plurality of command keywords and/or a wake-word for a Voice Assistant Service (VAS) and, based on the analysis, it is determined that the input sound data stream includes voice input data comprising a command keyword or a wake-word for a VAS. In response, the device takes appropriate action such as causing the media playback system to perform a command corresponding to the command keyword or sending at least part of the input sound data stream to the VAS.