Efficient Evaluation of Adversarial Robustness for Deep Hashing based Retrieval

Xunguang Wang, Xinyue Xu, Jiawang Bai, Xiaomeng Li

Keywords: Adversarial Attack, Adversarial Training, Deep Hashing, Similarity Retrieval

Abstract: Deep hashing has been extensively applied to massive image retrieval due to its efficiency and effectiveness. Recently, several adversarial attacks have been presented to reveal the vulnerability of deep hashing models against adversarial examples. However, existing attack methods suffer in degraded performance or inefficiency because they underutilize the semantic relations between original samples or spend a lot of time learning from these samples. In this paper, we propose a novel Pharos-guided Attack, dubbed \textbf{PgA}, to evaluate the adversarial robustness of deep hashing networks efficiently. Specifically, we design \textit{pharos code} to represent the semantics of the benign image, which preserves the similarity with semantically related samples and dissimilarity with irrelevant examples. It is proven that we can quickly calculate the pharos code via a simple math formula rather than time-consuming iterative procedures. Thus, PgA can directly conduct a reliable and efficient attack on deep hashing-based retrieval by maximizing the similarity between the hash code of the adversarial example and the pharos code. Extensive experiments on the benchmark datasets verify that the proposed algorithm outperforms the prior state-of-the-arts in both attack strength and speed.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics

Submission Guidelines: Yes

Please Choose The Closest Area That Your Submission Falls Into: Social Aspects of Machine Learning (eg, AI safety, fairness, privacy, interpretability, human-AI interaction, ethics)