Voice Assistant Request Surplus Detection via Audio Fingerprinting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Voice-activated electronic devices often experience errors, increased latency, or crashes when a large number of devices are triggered simultaneously by the same audio content, such as during media events, leading to overwhelming backend systems and poor user experiences.
Innovation Solution
Implementing sound profile recognition techniques on both the backend system and the devices to identify and ignore duplicate audio triggers, using audio fingerprinting to match sound profiles and prevent unnecessary processing, and employing request surplus protection mechanisms to manage excessive requests.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If voice-activated devices process all detected audio triggers simultaneously, then responsiveness to user commands is improved, but backend system stability deteriorates due to overwhelming request loads
Solution Approach 1:
The system performs preliminary sound profile recognition and duplicate detection before forwarding requests to the backend. By identifying and filtering duplicate audio triggers upfront using audio fingerprinting, the system prevents overwhelming the backend with redundant requests while maintaining rapid response to genuine user commands.
Solution Approach 2:
The patent introduces an intermediary layer (sound profile recognition module and request surplus protection mechanism) between the voice-activated devices and the backend system. This intermediary filters and manages incoming requests, blocking duplicates and regulating traffic flow to prevent system overload while allowing legitimate requests through.
2Reliability
If duplicate audio triggers are processed, then comprehensive command recognition is achieved, but processing latency increases due to redundant backend operations
Solution Approach 1:
The system extracts and removes duplicate audio triggers from the request stream using sound profile recognition and audio fingerprinting. By identifying matching sound profiles and filtering out redundant requests before they reach the backend, the system eliminates unnecessary processing delays while preserving genuine user commands.
Solution Approach 2:
The patent implements a mechanism to skip processing of duplicate requests by rapidly comparing audio fingerprints against known sound profiles. When a duplicate is detected, the system rushes through the filtering process and blocks the redundant request immediately, preventing it from entering the slower backend processing pipeline.
3Loss of information
If all detected audio requests are forwarded to the backend, then complete user intent analysis is possible, but energy consumption increases due to excessive data transmission and processing
Solution Approach 1:
The system performs preliminary filtering of audio requests using sound profile recognition and duplicate detection before transmission to the backend. By pre-processing requests locally and blocking obvious duplicates, the system reduces the volume of data that needs to be transmitted and processed remotely, thereby conserving energy while maintaining the ability to analyze genuine user intent.
Solution Approach 2:
The patent implements a mechanism to discard duplicate audio requests that have already been processed or are identical to previous requests. By using audio fingerprinting to identify and discard redundant requests, the system recovers energy that would otherwise be spent on transmitting and processing duplicate data, while preserving complete analysis of unique user commands.
Data Source
AI summary
Systems and methods for determining that artificial commands, in excess of a threshold value, are detected by multiple voice activated electronic devices is described herein. In some embodiments, numerous voice activated electronic devices may send audio data representing a phrase to a backend system at a substantially same time. Text data representing the phrase, and counts for instances of that text data, may be generated. If the number of counts exceeds a predefined threshold, the backend system may cause any remaining response generation functionality that particular command that is in excess of the predefined threshold to be stopped, and those devices returned to a sleep state. In some embodiments, a sound profile unique to the phrase that caused the excess of the predefined threshold may be generated such that future instances of the same phrase may be recognized prior to text data being generated, conserving the backend system's resources.