Dynamic Recognition of Speakers for Consent Management by Contrastive Embedding Replay
This addresses the practical problem of consent management for voice assistant users, though it appears incremental as it builds on existing speaker recognition methods.
The paper tackles the challenge of dynamic speaker registration, removal, and re-registration for consent management in voice assistants by proposing a system using contrastive embedding replay. The results show it outperforms existing approaches with improved memory efficiency and dynamic capabilities.
Voice assistants overhear conversations and a consent management mechanism is required. Consent management can be implemented using speaker recognition. Users that do not give consent enrol their voice and all their further recordings are discarded. Building speaker recognition-based consent management is challenging as dynamic registration, removal, and re-registration of speakers must be efficiently handled. This work proposes a consent management system addressing the aforementioned challenges. A contrastive based training is applied to learn the underlying speaker equivariance inductive bias. The contrastive features for buckets of speakers are trained a few steps into each iteration and act as replay buffers. These features are progressively selected using a multi-strided random sampler for classification. Moreover, new methods for dynamic registration using a portion of old utterances, removal, and re-registration of speakers are proposed. The results verify memory efficiency and dynamic capabilities of the proposed methods and outperform the existing approach from the literature.