Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2605.25663 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914599371014144 |
|---|---|
| author | Tariolle, Florent Yger, Florian |
| author_facet | Tariolle, Florent Yger, Florian |
| contents | 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 Target Selection (OTS), a lightweight wrapper that switches an untargeted attack to a targeted objective early in its trajectory, locking onto whichever non-true class currently leads. OTS requires no architectural modification to the underlying attack, no gradient access, and no a priori target-class knowledge.
We validate OTS on three score-based attacks (SimBA, Square Attack with cross-entropy loss, and Bandits) across five standard ImageNet classifiers (4,500 runs). On random-search attacks, OTS closely tracks oracle performance, with gains up to +27 pp in success rate and 43% relative reduction in censored-mean iterations on ResNet-50. On gradient-estimation attacks (Bandits) and attacks with margin loss, OTS is redundant, a negative result that reinforces our interpretation of OTS as a margin-loss surrogate. On adversarially-trained models, a bimodal difficulty distribution eliminates the regime where targeting helps. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_25663 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Opportunistic Target Selection: Early Directional Commitment for Query-Efficient Black-Box Adversarial Attacks Tariolle, Florent Yger, Florian Machine Learning Computer Vision and Pattern Recognition 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 Target Selection (OTS), a lightweight wrapper that switches an untargeted attack to a targeted objective early in its trajectory, locking onto whichever non-true class currently leads. OTS requires no architectural modification to the underlying attack, no gradient access, and no a priori target-class knowledge. We validate OTS on three score-based attacks (SimBA, Square Attack with cross-entropy loss, and Bandits) across five standard ImageNet classifiers (4,500 runs). On random-search attacks, OTS closely tracks oracle performance, with gains up to +27 pp in success rate and 43% relative reduction in censored-mean iterations on ResNet-50. On gradient-estimation attacks (Bandits) and attacks with margin loss, OTS is redundant, a negative result that reinforces our interpretation of OTS as a margin-loss surrogate. On adversarially-trained models, a bimodal difficulty distribution eliminates the regime where targeting helps. |
| title | Opportunistic Target Selection: Early Directional Commitment for Query-Efficient Black-Box Adversarial Attacks |
| topic | Machine Learning Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2605.25663 |