Contactless Voice Authentication System Using Cloud Speech Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional authentication systems are inefficient due to reliance on contact methods, lack of advanced security features, and compatibility issues across different devices and platforms, leading to security risks and inconvenience.
Innovation Solution
The implementation of a contactless voice recognition system that uses cloud services and ephemeral container instances for secure and cross-platform compatible authentication, allowing users to authenticate using various devices without physical contact or specialized hardware.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional authentication systems require contact with a device (manual entry or physical manipulation), then authentication can be performed, but the process becomes slow and inefficient
Solution Approach 1:
The patent replaces mechanical contact-based authentication (manual entry, physical manipulation of devices) with voice-based acoustic authentication. The system captures voice signals through a microphone and processes them through speech-to-text conversion and authentication algorithms, eliminating the need for physical contact while maintaining security and improving speed.
2Adaptability or versatility
If conventional systems use dedicated authentication devices (cards, keys, fobs), then authentication data can be stored, but users must carry multiple devices which is inconvenient
Solution Approach 1:
The patent extracts the authentication function from physical dedicated devices (cards, keys, fobs) and embeds it directly into the user's smartphone application. The voice authentication capability is integrated into the messaging app, eliminating the need to carry separate authentication devices while maintaining versatile authentication methods.
Solution Approach 2:
The patent creates a universal authentication system where a single smartphone application can perform multiple functions: messaging, voice authentication, and transaction authorization. The voice biometric authentication works across different contexts (personal messages, business transactions) without requiring different devices, providing multi-functionality within a single platform.
3Reliability
If conventional systems rely on authentication data from dedicated devices, then authentication can be performed, but the system is prone to security risks from stolen data or devices
Solution Approach 1:
The patent replaces physical authentication devices that can be stolen with voice biometric authentication that is inherently more secure. Voice patterns are difficult to steal or replicate compared to physical devices, and the system can detect spoofing attempts. The authentication is tied to the user's biological characteristics rather than a physical object that can be lost or stolen.
Solution Approach 2:
The system performs preliminary voice pattern analysis and establishes baseline voice profiles during setup. This preliminary action creates a reference framework that enables real-time authentication decisions, allowing the system to quickly verify users while maintaining high security standards against theft and spoofing.
4Reliability
If conventional systems do not employ advanced authentication methods (facial recognition, voice recognition), then computational resources are saved, but advanced security features are unavailable
Solution Approach 1:
The patent uses speech-to-text conversion as an intermediary step that transforms complex voice signal processing into manageable text data for authentication. This intermediary approach simplifies the computational requirements while maintaining the security benefits of voice biometrics, making advanced authentication feasible on mobile devices.
Data Source
AI summary
Systems and methods for contactless authorization using voice recognition is disclosed. The system may include one or more memory units storing instructions and one or more processors configured to execute the instructions to perform operations. The operations may include receiving user data comprising a user identifier, an audio data having a first data format, and a client device identifier. The operations may include generating a processed audio data based on the received audio data. The processed audio data may have a second data format. The operations may include transmitting, to a speech module, the processed audio data. The operations may include receiving from the speech module, a voice match result. In some embodiments, the operations include authenticating a user based on the voice match result and transmitting, to a client device associated with the client device identifier, a client notification comprising a result of the authentication.