RO-SVD: A Reconfigurable Hardware Copyright Protection Framework for AIGC Applications

Zhuoheng Ran,Muhammad A.A. Abdelgawad,Zekai Zhang,Ray C.C. Cheung,Hong Yan

2024-06-17

Abstract:The dramatic surge in the utilisation of generative artificial intelligence (GenAI) underscores the need for a secure and efficient mechanism to responsibly manage, use and disseminate multi-dimensional data generated by artificial intelligence (AI). In this paper, we propose a blockchain-based copyright traceability framework called ring oscillator-singular value decomposition (RO-SVD), which introduces decomposition computing to approximate low-rank matrices generated from hardware entropy sources and establishes an AI-generated content (AIGC) copyright traceability mechanism at the device level. By leveraging the parallelism and reconfigurability of field-programmable gate arrays (FPGAs), our framework can be easily constructed on existing AI-accelerated devices and provide a low-cost solution to emerging copyright issues of AIGC. We developed a hardware-software (HW/SW) co-design prototype based on comprehensive analysis and on-board experiments with multiple AI-applicable FPGAs. Using AI-generated images as a case study, our framework demonstrated effectiveness and emphasised customisation, unpredictability, efficiency, management and reconfigurability. To the best of our knowledge, this is the first practical hardware study discussing and implementing copyright traceability specifically for AI-generated content.

Distributed, Parallel, and Cluster Computing,Computer Vision and Pattern Recognition

What problem does this paper attempt to address?

The problem this paper attempts to address is the copyright protection of AI-generated content (AIGC). With the widespread application of generative artificial intelligence (GenAI), AI-generated content is playing an increasingly important role in the creative industry, but it also raises widespread copyright disputes. These issues include how to effectively manage and track AI-generated content to prevent unauthorized use and distribution.

Specifically, the paper proposes a blockchain-based copyright tracing framework called Ring Oscillator-Singular Value Decomposition (RO-SVD). This framework utilizes low-rank matrices generated by hardware entropy sources and employs Singular Value Decomposition (SVD) technology to extract and manage these matrices, thereby providing a device-level copyright tracing mechanism for AI-generated content. The main contributions of the paper include:

1. **Low-rank characteristics of hardware entropy sources**: The paper demonstrates that matrices generated by hardware entropy sources have approximately low-rank characteristics, which makes copyright tracing of AI-generated content possible.
2. **RO-SVD framework**: A framework named RO-SVD is proposed, which can automatically and efficiently tag and manage AI-generated content and achieve copyright protection based on blockchain technology.
3. **SVD computational component**: Singular Value Decomposition (SVD) is introduced as a powerful computational component to separate matrix primitives generated by FPGA and serve as a synergistic design of Physical Unclonable Functions (PUF) and True Random Number Generators (TRNG) to manage and tag a large amount of content.
4. **Hardware-software co-design prototype**: A hardware-software co-design prototype is developed, which can be easily built on existing AI acceleration devices.

Through these contributions, the paper provides a practical hardware solution for the copyright tracing of AI-generated content, particularly in terms of efficiency, cost, management, unpredictability, and reconfigurability.