Work

Two Heads Are Better Than One: Averaging along Fine-Tuning to Improve Targeted Transferability

Year: 2025

Type: conference-paper

Abstract: With much longer optimization time than that of untargeted attacks notwithstanding, the transferability of targeted attacks is still far from satisfactory. Recent studies reveal that fine-tuning an existing adversarial example (AE) in feature space can efficiently boost its targeted transferability. However, existing fine-tuning schemes only utilize the endpoint and ignore the valuable information in the fine-tuning trajectory. Noting that the vanilla fine-tuning trajectory tends to oscillate around the periphery of a flat region of the loss surface, we propose averaging over the fine-tuning trajectory to pull the crafted AE towards a more centered region. We compare the proposed method with existing fine-tuning schemes by integrating them with state-of-the-art targeted attacks in various attacking scenarios. Experimental results uphold the superiority of the proposed method in boosting targeted transferability. The code is available at github.com/zengh5/Avg_FT.

Source: IEEE International Conference on Acoustics Speech and Signal Processing

DOI: 10.1109/icassp49660.2025.10890045

Authors: Hui Zeng, Sanshuai Cui, Biwei Chen, Anjie Peng

Institutions: Southwest University of Science and Technology, City University of Macau, Beijing Normal University, Zhuhai

Language: English

Topic: Adversarial Robustness in Machine Learning

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: green

Funders: Beijing Normal University, Guangdong Province Key Laboratory of Information Security Technology

Award: G3835580601