Location-Based Voice Authentication via Spatial Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Voice-controlled devices face challenges in distinguishing between multiple users, particularly in environments with multiple users and when users move, leading to issues with associating audio signals with the correct user and verifying user identity through voice recognition, as recorded or synthesized voices can spoof authentication phrases.
Innovation Solution
A location-based voice recognition system that correlates the source location of audio signals with the user's physical location to authenticate users, using a combination of acoustic localization and user localization techniques to determine the source location and user location, ensuring that only authorized users can access services.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If voice recognition algorithms are used for user authorization, then user authentication can be performed without physical input, but voice authentication phrases can be easily spoofed by recorded or synthesized inputs
Solution Approach 1:
The patent introduces location information as an intermediary verification factor between the voice authentication system and the user. By correlating the source location of the audio signal with the user's known location (from mobile device GPS or check-in data), the system adds a spatial dimension to authentication that cannot be spoofed by voice alone, thus maintaining ease of voice operation while improving security reliability
Solution Approach 2:
The patent transitions from one-dimensional voice biometric authentication to two-dimensional authentication by adding spatial location correlation. This dimensional expansion requires that not only must the voice match the user's biometric profile, but the audio signal must also originate from the user's expected physical location, preventing remote spoofing attacks
2Adaptability or versatility
If voice-controlled devices process audio signals from multiple users, then the device can serve multiple users, but it becomes difficult to distinguish which user is providing a voice command and associate audio signals with the correct user
Solution Approach 1:
The patent applies local quality by associating specific location data with each user's voice commands. Instead of treating all audio signals uniformly, the system correlates each audio signal's source location with the known locations of specific users (obtained from their mobile devices), enabling precise identification of which user is speaking even in multi-user environments
Solution Approach 2:
The system uses feedback from mobile devices that provide users' current locations to the voice-controlled device. This continuous location feedback allows the system to dynamically update which user is expected to be at the audio signal source, improving user identification accuracy as users move through different locations
Data Source
AI summary
Systems and methods for providing location based voice recognition include receiving, through a first microphone, an audio signal from a first user that includes an audio command requesting a service that requires user authorization before access to at least a portion of the service is granted. The user authorization is based on voice recognition (e.g., voice authentication and/or voice identification) of the audio signal. The source location of the audio signal is determined and a user location of the first user is determined. If the source location of the audio signal correlates with the user location, voice recognition on the audio signal may be performed. The first user may be authorized to access the service based on the voice recognition performed on the audio signal.