PulseAugur
EN
LIVE 11:30:17

New method boosts black-box adversarial attack efficiency

By PulseAugur Editorial · [2 sources] ·

Researchers have developed a new method called Opportunistic Target Selection (OTS) to improve the efficiency of black-box adversarial attacks. OTS acts as a wrapper, allowing attacks to switch from an untargeted objective to a targeted one early in their process, thereby reducing wasted queries. This technique requires no modifications to the underlying attack model or access to gradients, and it was validated on ImageNet classifiers, showing significant improvements in success rate and query efficiency. AI

IMPACT Enhances the efficiency of adversarial attacks, potentially impacting AI model robustness testing and security.

RANK_REASON The cluster contains an academic paper detailing a new research method in adversarial attacks.

Read on arXiv cs.LG →

AI-generated summary · Google Gemini · from 2 sources. How we write summaries →

New method boosts black-box adversarial attack efficiency

COVERAGE [2]

  1. arXiv cs.LG TIER_1 English(EN) · Florent Tariolle, Florian Yger ·

    Opportunistic Target Selection: Early Directional Commitment for Query-Efficient Black-Box Adversarial Attacks

    arXiv:2605.25663v1 Announce Type: new Abstract: Black-box adversarial attacks that minimize only the ground-truth confidence suffer from class drift: perturbations wander through the feature space without committing to a specific adversarial class, wasting queries on diffuse, und…

  2. arXiv cs.LG TIER_1 English(EN) · Florian Yger ·

    Opportunistic Target Selection: Early Directional Commitment for Query-Efficient Black-Box Adversarial Attacks

    Black-box adversarial attacks that minimize only the ground-truth confidence suffer from class drift: perturbations wander through the feature space without committing to a specific adversarial class, wasting queries on diffuse, undirected progress. We introduce Opportunistic Tar…

RELATED ENTITIES

RELATED TOPICS