Voice-Activated Device False Trigger Prevention via Audio Watermarking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Voice-activated devices often receive false triggers from unintended sources, leading to unintended activation, such as during playback of television commercials, causing them to execute unwanted commands.
Innovation Solution
Inserting signal markers into content assets at locations corresponding to triggers, which are inaudible to users but detectable by voice-activated devices, to prevent false triggers by instructing the device to ignore voice commands associated with these triggers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If voice-activated devices continuously monitor for trigger words to enable hands-free operation, then ease of operation is improved, but reliability deteriorates due to false triggers from unintended sources such as television commercials
Solution Approach 1:
The patent introduces an audio watermark as an intermediary signal that is embedded in legitimate voice commands but absent in false triggers from television commercials. The voice-activated device detects this watermark to verify the authenticity of the trigger source, thereby resolving the contradiction between maintaining hands-free operation and preventing false activations
Solution Approach 2:
The system performs preliminary verification by checking for the presence of an audio watermark before executing voice commands. This preliminary action occurs at the detection stage, allowing the device to distinguish between legitimate and spurious triggers before committing to command execution, thus improving reliability while preserving ease of operation
Data Source
AI summary
A computing device may be configured to receive a content asset and to determine whether the content asset comprises one or more triggers. The trigger may be a word, phrase, or passcode that alerts a voice activated device to the presence of a voice command and may serve as an instruction to the voice activated device to cause execution of the voice command. In response to determining that the content asset comprises one or more triggers, the computing device may be configured to insert one or more signal markers into the content asset at a location corresponding to the one or more triggers, and to cause transmission and/or presentation of the content asset with the one or more signal markers. The signal markers may cause a voice activated device to ignore a voice command in the content asset, despite the presence of one or more triggers.