Dare to Plagiarize? Plagiarized Painting Recognition and Retrieval

2025-06-29

AI Summary

This work addresses the dual challenges of artwork plagiarism detection and interpretable retrieval. We propose a visual similarity analysis framework integrating generative AI and metric learning. First, we construct the first synthetic dataset simulating artist-style plagiarism using generative AI—enabling controlled evaluation of plagiarism detection. Second, we extract image features using DINOv2, achieving 97.2% plagiarism classification accuracy with fixed, non-learned features, thereby validating the sufficiency of foundational visual representations. Third, we fine-tune the feature extractor via metric learning, preserving strong discriminability while improving mean average precision (mAP) for retrieval by 12 percentage points to 41.0%, exposing an intrinsic trade-off between discrimination and retrieval performance. This study is the first to systematically investigate: (i) generative modeling of AI-simulated plagiarism, (ii) joint optimization of discrimination and retrieval, and (iii) the applicability boundary of vision foundation models in art copyright protection.

Technology Category

• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Machine Learning: Adversarial Learning & Robustness
• Humans and AI: Other Foundations of Human Computation & AI

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Web Mining and Content Analysis: Web data generation and simulation
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Art plagiarism detection plays a crucial role in protecting artists' copyrights and intellectual property, yet it remains a challenging problem in forensic analysis. In this paper, we address the task of recognizing plagiarized paintings and explaining the detected plagarisms by retrieving visually similar authentic artworks. To support this study, we construct a dataset by collecting painting photos and synthesizing plagiarized versions using generative AI, tailored to specific artists' styles. We first establish a baseline approach using off-the-shelf features from the visual foundation model DINOv2 to retrieve the most similar images in the database and classify plagiarism based on a similarity threshold. Surprisingly, this non-learned method achieves a high recognition accuracy of 97.2% but suffers from low retrieval precision 29.0% average precision (AP). To improve retrieval quality, we finetune DINOv2 with a metric learning loss using positive and negative sample pairs sampled in the database. The finetuned model greatly improves retrieval performance by 12% AP over the baseline, though it unexpectedly results in a lower recognition accuracy (92.7%). We conclude with insightful discussions and outline directions for future research.

Problem

• Detect plagiarized paintings to protect artists' copyrights
• Retrieve visually similar authentic artworks for plagiarism explanation
• Improve retrieval precision while maintaining high recognition accuracy

Innovation

• Uses generative AI to synthesize plagiarized paintings
• Fine-tunes DINOv2 with metric learning for retrieval
• Achieves high accuracy with non-learned similarity threshold