CVDec 12, 2023

Collapse-Aware Triplet Decoupling for Adversarially Robust Image Retrieval

arXiv:2312.07364v44 citationsh-index: 18Has CodeICML
Originality Highly original
AI Analysis

This work addresses adversarial robustness in image retrieval, an incremental improvement for deep metric learning applications.

The paper tackles the limitations of weak adversaries and model collapse in adversarially robust image retrieval by proposing CA-TRIDE, which outperforms existing defense methods on three datasets.

Adversarial training has achieved substantial performance in defending image retrieval against adversarial examples. However, existing studies in deep metric learning (DML) still suffer from two major limitations: weak adversary and model collapse. In this paper, we address these two limitations by proposing Collapse-Aware TRIplet DEcoupling (CA-TRIDE). Specifically, TRIDE yields a stronger adversary by spatially decoupling the perturbation targets into the anchor and the other candidates. Furthermore, CA prevents the consequential model collapse, based on a novel metric, collapseness, which is incorporated into the optimization of perturbation. We also identify two drawbacks of the existing robustness metric in image retrieval and propose a new metric for a more reasonable robustness evaluation. Extensive experiments on three datasets demonstrate that CA-TRIDE outperforms existing defense methods in both conventional and new metrics. Codes are available at https://github.com/michaeltian108/CA-TRIDE.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes