WAKE: Watermarking Audio with Key Enrichment

TL;DR

WAKE is a novel key-controlled audio watermarking framework that enhances security, supports multiple embeddings, and handles variable-length watermarks, outperforming existing methods in quality and accuracy.

Contribution

It introduces the first key-enriched, controllable audio watermarking framework that addresses security, multiple embedding, and variable-length watermark challenges.

Findings

Outperforms existing models in watermarked audio quality.

Achieves higher watermark detection accuracy.

Supports multiple embeddings and variable-length watermarks.

Abstract

As deep learning advances in audio generation, challenges in audio security and copyright protection highlight the need for robust audio watermarking. Recent neural network-based methods have made progress but still face three main issues: preventing unauthorized access, decoding initial watermarks after multiple embeddings, and embedding varying lengths of watermarks. To address these issues, we propose WAKE, the first key-controllable audio watermark framework. WAKE embeds watermarks using specific keys and recovers them with corresponding keys, enhancing security by making incorrect key decoding impossible. It also resolves the overwriting issue by allowing watermark decoding after multiple embeddings and supports variable-length watermark insertion. WAKE outperforms existing models in both watermarked audio quality and watermark detection accuracy. Code, more results, and demo page: https://thuhcsi.github.io/WAKE.