Voice Bot Synthetic Responses for Voice Privacy Protection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies lack regulations to prevent misuse of voice recordings and fail to protect user privacy in scenarios where voice data is collected without consent, particularly in unsolicited calls and spam calls that can record a person's voice without permission.
Innovation Solution
A device is configured to allow users to select and render text-based labels for synthesized voice responses, using synthetic speech generators to respond to calls without exposing their actual voice, with features like auto-muting the microphone for unknown callers and providing ranked or user-specified message lists.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If users respond to calls using their own voice, then communication effectiveness is improved, but voice privacy and security are compromised due to unauthorized recording
Solution Approach 1:
The patent creates a synthetic copy of the user's voice that mimics their vocal characteristics without using their actual voice recordings. The system generates artificial voice responses that sound like the user but are completely synthesized, allowing communication while preventing exposure of the user's real voice data to callers or third parties.
Solution Approach 2:
The patent introduces a synthetic voice generation system as an intermediary between the user and the caller. Instead of the user's actual voice being transmitted, the system acts as a mediator that converts text inputs into synthetic speech that resembles the user's voice, thereby protecting the user's voice privacy while maintaining communication flow.
2Object-affected harmful factors
If synthesized voice responses are used to protect privacy, then voice recording security is improved, but communication naturalness may deteriorate
Solution Approach 1:
The patent adjusts various parameters of the synthetic voice generation including pitch, tone, speech rate, and vocal timbre to match the user's natural speaking characteristics. By fine-tuning these acoustic parameters, the system creates synthetic speech that closely resembles the user's natural voice, making the communication feel more authentic and natural despite being synthesized.
3Loss of energy
If text-based synthesized responses are implemented, then bandwidth usage is reduced, but response customization flexibility is limited
Solution Approach 1:
The patent implements a universal text-to-speech system that can generate multiple types of responses (polite refusals, busy messages, do-not-disturb notifications, etc.) from a single text-based interface. The system handles various communication scenarios using the same synthetic voice generation mechanism, providing diverse response options while maintaining efficient bandwidth usage through text-based input rather than audio recording.
Data Source
AI summary
A device is configured to communicate on a mobile communications network. An incoming call is received, and it is determined that the incoming call meets a predetermined criteria indicating a probable source of the incoming call. On a display of the device, an option is rendered for answering the incoming call with a generated voice response in lieu of a voice of a user of the device. Text options for generating a voice response are also rendered. The incoming call is answered and generated speech corresponding to the selected text option is sent.