Speaker-Identified Privacy Mode in Speech Processing Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing speech processing systems do not effectively manage privacy modes associated with specific user identities, leading to unintended personalization issues when privacy modes are not properly deactivated.
Innovation Solution
A speech processing system that configures and executes a privacy mode based on a user's unique identifier, allowing users to activate privacy mode through voice commands and ensuring that data related to user utterances is not stored long-term unless privacy mode is deactivated.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the speech processing system stores data related to user utterances for long-term personalization, then the system can provide better personalized interactions, but user privacy is compromised when privacy mode should be active
Solution Approach 1:
The system dynamically adjusts its data storage behavior based on the detected speaker identity and their associated privacy mode settings. When a user in privacy mode speaks, the system immediately deletes their utterance data after processing, whereas non-privacy users have their data retained for personalization. This dynamic adaptation resolves the contradiction by making the system's personalization capability conditional rather than universal.
Solution Approach 2:
The system applies different data retention policies to different users based on their individual privacy mode configurations. Instead of a uniform approach, each user's data is treated differently according to their specific privacy preferences stored in the system. This local differentiation allows personalized interactions for some users while protecting privacy for others.
2Adaptability or versatility
If the system deactivates privacy mode for all users by default to enable personalization, then personalized interactions are available to everyone, but users who want privacy cannot obtain it
Solution Approach 1:
The system automatically detects the speaker's identity and applies the appropriate privacy mode setting without requiring manual user intervention. Users who have previously configured privacy preferences have their settings automatically applied when they speak, eliminating the need for them to manually activate or deactivate privacy mode each time.
Solution Approach 2:
The system uses speaker identification technology to provide feedback about which user is speaking, then automatically adjusts data storage behavior based on that user's privacy mode configuration. This closed-loop feedback mechanism ensures the correct privacy policy is applied to each user without requiring their active participation.
3Object-affected harmful factors
If the system implements speaker-specific privacy mode management, then user privacy is respected, but the system complexity increases due to additional speaker identification and mode management requirements
Solution Approach 1:
The existing speech recognition system is extended to also perform speaker identification and privacy mode management functions. Rather than adding completely separate systems, the patent makes the speech processing system multi-functional by integrating speaker verification and privacy policy enforcement into the same architecture that already handles speech-to-text conversion.
Data Source
AI summary
Techniques for configuring a speech processing system with a privacy mode that is associated with the identity of a user that activated the privacy mode are described. A user may speak an indication to have the speech processing system activate a privacy mode. When such an indication is detected by the speech processing system, the speech processing system determines an identity of the user, determines a unique system identifier associated with the user, and generates a privacy mode flag. The speech processing system then associates the privacy mode flag with the user's unique system identifier. The privacy mode flag indicates to components of the speech processing system that any data related to processing of the user's utterances should not be sent to long term storage, thus causing various components of the system to delete data once the respective component is finished processing with respect to an utterance of the user.