mlsanityData to 14 Oct 2026

LOCKEY: A Novel Approach to Model Authentication and Deepfake Tracking

arXiv 2409.07743 ↗First posted 12 Sep 2024cs.CRcited by 0

On the map: LLM safety and security → Watermarking generated content and models

Abstract (arXiv)

This paper presents a novel approach to deter unauthorized deepfakes and enable user tracking in generative models, even when the user has full access to the model parameters, by integrating key-based model authentication with watermarking techniques. Our method involves providing users with model parameters accompanied by a unique, user-specific key. During inference, the model is conditioned upon the key along with the standard input. A valid key results in the expected output, while an invalid key triggers a degraded output, thereby enforcing key-based model authentication. For user tracking, the model embeds the user's unique key as a watermark within the generated content, facilitating the identification of the user's ID. We demonstrate the effectiveness of our approach on two types of models, audio codecs and vocoders, utilizing the SilentCipher watermarking method. Additionally, we assess the robustness of the embedded watermarks against various distortions, validating their reliability in various scenarios.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Watermarking generated content and models, one of 12 directions in LLM safety and security. Builds on 18 of its 20 references that are in the map. First posted on arXiv 12 Sep 2024.

Builds on

18 of 20 references.

Closest ancestors · 1

Same direction, newest first

  1. 2023A Watermark for Large Language Models24 Jan

Other references · 17

  1. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  2. 2020HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesiscited by 1,154
  3. 2022High Fidelity Neural Audio Compressioncited by 687
  4. 2023Towards Universal Fake Image Detectors that Generalize Across Generative Modelscited by 490
  5. 2023Simple and Controllable Music Generationcited by 488
  6. 2023Consistency Trajectory Models: Learning Probability Flow ODE Trajectory of Diffusioncited by 459
  7. 2023DIRE for Diffusion-Generated Image Detectioncited by 390
  8. 2024Fast Timing-Conditioned Latent Audio Diffusioncited by 181
  9. 2024Proactive Detection of Voice Cloning with Localized Watermarkingcited by 78
  10. 2023WavMark: Watermarking for Audio Generationcited by 50
  11. 2023Level Up the Deepfake Detection: a Method to Effectively Discriminate Images Generated by GAN Architectures and Diffusion Modelscited by 21
  12. 2024SilentCipher: Deep Audio Watermarkingcited by 21
5 more
  1. 2024Latent Watermarking of Audio Generative Modelscited by 16
  2. 2023Collaborative Watermarking for Adversarial Speech Synthesiscited by 7
  3. 2022Robust One-Shot Singing Voice Conversioncited by 6
  4. 2021Source Mixing and Separation Robust Audio Steganographycited by 3
  5. 2023Iteratively Improving Speech Recognition and Voice Conversioncited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025DeepTracer: Tracing Stolen Model via Deep Coupled Watermarks
  2. 2023Performance-lossless Black-box Model Watermarking
  3. 2025Mitigating Watermark Forgery in Generative Models via Randomized Key Selection
  4. 2025Visual Watermarking in the Era of Diffusion Models: Advances and Challenges
  5. 2025On-Device Watermarking: A Socio-Technical Imperative For Authenticity In The Age of Generative AI
  6. 2023Invisible Watermarking for Audio Generation Diffusion Models
  7. 2023Towards Robust Model Watermark via Reducing Parametric Vulnerability
  8. 2024GROOT: Generating Robust Watermark for Diffusion-Model-Based Audio Synthesis