Semantic to Structure: Learning Structural Representations for Infringement Detection

2025-02-11

AI Summary

This paper addresses the novel copyright risk of “structural infringement”—where AI-generated content imitates the compositional structure of original works—by proposing the first end-to-end detection framework. Methodologically, it formally defines structural infringement; introduces SIA (Synthetic) and SIR (Real), the first benchmark dataset jointly annotated for semantic and structural attributes; designs a structured data synthesis strategy integrating diffusion models and large language models; and develops a detection method based on structural representation learning and contrastive learning. Experimental results demonstrate that the proposed approach significantly outperforms existing baselines on the SIA/SIR test sets, validating the critical role of structural representations in infringement identification. The work establishes a practical technical pathway and an evaluation paradigm for AIGC copyright protection, advancing both methodology and empirical assessment in generative AI governance.

Technology Category

• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Philosophy and Ethics of AI: Artificial General Intelligence
• Natural Language Processing: Generation

Application Category

• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• Web Mining and Content Analysis: Web data quality in the era of algorithmically-generated content

Abstract

Structural information in images is crucial for aesthetic assessment, and it is widely recognized in the artistic field that imitating the structure of other works significantly infringes on creators' rights. The advancement of diffusion models has led to AI-generated content imitating artists' structural creations, yet effective detection methods are still lacking. In this paper, we define this phenomenon as"structural infringement"and propose a corresponding detection method. Additionally, we develop quantitative metrics and create manually annotated datasets for evaluation: the SIA dataset of synthesized data, and the SIR dataset of real data. Due to the current lack of datasets for structural infringement detection, we propose a new data synthesis strategy based on diffusion models and LLM, successfully training a structural infringement detection model. Experimental results show that our method can successfully detect structural infringements and achieve notable improvements on annotated test sets.

Problem

• Detect structural infringement in images
• Develop datasets for infringement detection
• Propose AI-based detection and synthesis methods

Innovation

• Diffusion models for data synthesis
• LLM-based structural infringement detection
• Manually annotated datasets for evaluation