Playback Device Gatekeeping for Local Voice Intent Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing voice-assisted media playback systems require a wake word to invoke a voice assistant service, leading to slower response times and potential privacy concerns due to the need for cloud-based processing of voice inputs.
Innovation Solution
Implementing a command-keyword engine within network microphone devices that detects and processes voice commands locally, without a wake word, using a local natural language unit to recognize a limited library of keywords and apply confirmation conditions to reduce false positives.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a wake word is required to invoke voice assistant service, then the system can filter out unwanted voice inputs, but the response time increases and privacy concerns arise due to cloud-based processing
Solution Approach 1:
The patent segments voice processing into two distinct paths: a local keyword spotting engine that handles specific commands immediately, and a cloud-based voice assistant for general queries. This segmentation allows frequently used commands to be processed locally without wake word delays, while maintaining the filtering benefits for other inputs.
Solution Approach 2:
The patent introduces a local keyword spotting engine as an intermediary between the microphone and the cloud-based voice assistant. This intermediary captures and processes specific keywords locally, preventing them from being sent to the cloud, thereby reducing response time and addressing privacy concerns while maintaining accurate filtering.
2Adaptability or versatility
If cloud-based processing is used for all voice inputs, then comprehensive voice assistant functionality is achieved, but privacy concerns increase and response time decreases
Solution Approach 1:
The patent extracts specific keyword processing functionality from the cloud-based voice assistant and implements it locally in the network microphone device. This extraction allows common commands to be processed immediately without cloud communication delays, while the voice assistant retains comprehensive functionality for non-routine queries.
3Adaptability or versatility
If cloud-based processing is used for all voice inputs, then comprehensive voice assistant functionality is achieved, but privacy concerns increase due to transmission of all voice data
Solution Approach 1:
The patent extracts specific keyword processing from cloud-based handling and performs it locally. This prevents transmission of routine command data to the cloud, reducing privacy concerns while maintaining comprehensive voice assistant functionality for more complex or non-routine queries.
Solution Approach 2:
The patent implements local keyword spotting capability within the network microphone device, creating a quality distinction between locally processed commands and cloud-based assistant functions. This local processing capability addresses privacy concerns by keeping sensitive command data on-device while preserving comprehensive functionality through cloud integration.
4Speed
If keyword detection is implemented without secondary confirmation, then response speed increases, but false positives increase reducing reliability
Solution Approach 1:
The patent applies preliminary confirmation conditions to keyword detection before triggering commands. These conditions evaluate contextual factors such as voice characteristics, playback state, and environmental noise levels in advance, filtering out potential false positives before they execute commands, thus maintaining reliability without significantly impacting response speed.
Data Source
AI summary
In one aspect, an audio playback device having at least one microphone captures a voice input. The playback device detects, within the voice input, at least one keyword from among a plurality of command keywords supported by the playback device. The playback device determines, via a local natural language unit (NLU), an intent based on the keyword. The keyword is then evaluated based at least in part on a volume characteristic of the voice input. Based on the evaluation, the playback device either forgoes further processing of the voice input or performs a command in accordance with the determined intent.


