Voice Watermarking in Calls for Speaker Authenticity Verification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
AI-generated voices are used in call attacks to impersonate legitimate callers, spread misinformation, and commit fraud, particularly targeting vulnerable individuals in robocalls.
Innovation Solution
A communication session is configured to verify a voice signal with a watermarking signal, using a token-based system where a communication device generates and transmits a watermarking signal along with the voice signal, which is decoded by the receiving device to verify the authenticity of the speaker.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If AI-generated voice is used for communication, then communication convenience is improved, but security and authenticity deteriorate due to impersonation and fraud
Solution Approach 1:
A watermarking signal is introduced as an intermediary element between the voice signal and the receiver. This watermarking signal carries authentication information that verifies the identity and authenticity of the voice speaker, allowing the system to distinguish between legitimate and fraudulent AI-generated voices without affecting communication convenience
Solution Approach 2:
The system implements a feedback mechanism where the transmitter embeds authentication information in the watermarking signal, and the receiver verifies this information to confirm voice authenticity. This feedback loop ensures that only verified voices are accepted, addressing the security concerns while maintaining ease of communication
2Reliability
If watermarking signal is added to verify voice authenticity, then reliability is improved, but device complexity increases
Solution Approach 1:
The watermarking signal is merged with the voice signal through modulation techniques, combining two functions (communication and authentication) into a single transmitted signal. This reduces the number of separate components needed and simplifies the overall system architecture while maintaining verification capability
Solution Approach 2:
The communication device is designed to perform multiple functions: it transmits voice information, embeds authentication watermarks, and processes verification signals. This multi-functionality reduces the need for separate dedicated components for each function, thereby reducing overall device complexity
Data Source
AI summary
Aspects of the subject disclosure may include, for example, negotiating a communication session with a second communication device over a communication network. The communication session is configured to verify a voice signal with a watermarking signal. Further embodiments include obtaining a token from a network device, generating the watermarking signal based on the token, and transmitting the voice signal and the watermarking signal to the second communication device. Other embodiments are disclosed.