Scholarlink

With the growth of decentralized finance (DeFi), non-fungible tokens (NFTs), particularly those based on artistic images, have become mainstream. However, copyright issues are increasingly severe, as forged or plagiarized artworks are minted and resold as NFTs, infringing creators’ intellectual property rights. Despite progress in NFT copyright protection, detection remains post-minting, and the rise of AI-generated models has intensified image forgery and style plagiarism, which existing methods fail to address. We propose ZKP-StylePatch, comprising: (1) a model to detect the origin of artistic images, and (2) a zero-knowledge proof mechanism to verify model output integrity and authenticity. Using GenImage and a style-transferred NFT dataset, experiments show ZKP-StylePatch achieves 91.35% accuracy in detecting artwork authenticity.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

ZKP-StylePatch: Hybrid NFT Anti-counterfeit Framework

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Tiantian Wu,
  • Yixuan Shen,
  • Fan Zhang,
  • You Jiang,
  • Lei Yao,
  • Bin Zhou,
  • Junbin Fang

摘要

With the growth of decentralized finance (DeFi), non-fungible tokens (NFTs), particularly those based on artistic images, have become mainstream. However, copyright issues are increasingly severe, as forged or plagiarized artworks are minted and resold as NFTs, infringing creators’ intellectual property rights. Despite progress in NFT copyright protection, detection remains post-minting, and the rise of AI-generated models has intensified image forgery and style plagiarism, which existing methods fail to address. We propose ZKP-StylePatch, comprising: (1) a model to detect the origin of artistic images, and (2) a zero-knowledge proof mechanism to verify model output integrity and authenticity. Using GenImage and a style-transferred NFT dataset, experiments show ZKP-StylePatch achieves 91.35% accuracy in detecting artwork authenticity.