欢迎访问《图学学报》 分享到：
图学学报

图学学报 ›› 2022, Vol. 43 ›› Issue (1): 93-100.DOI: 10.11996/JG.j.2095-302X.2022010093

• 计算机图形学与虚拟现实 • 上一篇    下一篇

基于多视图网络三维形状检索的通用扰动攻击

  

  1. 广州大学计算机科学与网络工程学院，广东 广州 510006
  • 出版日期:2022-02-28 发布日期:2022-02-16
  • 基金资助:
    国家自然科学基金项目(62072126，61772164)；广州市科技计划项目(202002030263，202102010419)，广州大学校内科研人才培育项目 (XJ2021001901) 

MvUPA: universal perturbation attack against 3D shape retrieval based on multi-view networks

  1. School of Computer Science and Cyber Engineering, Guangzhou University, Guangzhou Guangdong 510006, China
  • Online:2022-02-28 Published:2022-02-16
  • Supported by:
    National Natural Science Foundation of China (62072126, 61772164); Guangzhou Science and Technology Plan Project (202002030263, 202102010419); Research Talents Cultivation Project of Guangzhou University (XJ2021001901) 

PDF (PC)

89

摘要： 几何深度学习模型在三维形状检索任务中已应用，其安全评估工作也引起了研究者们的关注。 该文针对三维形状检索评估提出一种基于多视图通用扰动攻击(MvUPA)的对抗攻击方法，其具有高成功率的攻 击效果。首先设计多视角深度全景图检索模型，训练适用于视图类三维形状检索的高效嵌入向量；其次，为三 维形状检索提出有益于通用扰动更新的损失函数方案和攻击机制。该损失函数方案同时融合了三元损失和标签 损失，提升了对相近拓扑异类样本和差异拓扑同类样本的对抗扰动生成。通过实验验证了 MvUPA 在多个视图 类检索模型上攻击的有效性和稳定性，攻击指标下降率(DR)最高达 94.52%；融合损失函数相比单个损失函数 DR 指标提高约 3.0%~5.5%。

关键词: 三维形状检索, 多视图通用扰动攻击, 通用扰动攻击, 几何深度学习, 融合损失

Abstract: Geometric deep learning models have been applied in 3D shape retrieval task successfully, and their security evaluation is also drawing the attention of researchers. This paper proposed a method of multi-view universal perturbation attack (MvUPA) for 3D shape retrieval evaluation, so as to generate perturbed samples with a higher success rate of attack. Firstly, a multi-view depth panoramic map-based network was designed to train an efficient embedding representation for multi-view 3D shape retrieval. Secondly, a fusion loss function and its attack mechanism beneficial to multi-input UPA was proposed. The loss function combined triplet loss and label loss, thereby improving the perturbation generation for different categories of samples with similar topology and same category samples with different topology. The experiments validated the attack effectiveness and stability of MvUPA on multi-view retrieval models. MvUPA brought the decrease rate (DR) up to 94.52%, and the DR of the fused loss function was about 3.0%–5.5% higher than that of a single loss function.  

Key words: 3D shape retrieval, multi-view universal perturbation attack, universal perturbation attack, geometric deep learning, fusion loss

中图分类号: