Making an Invisibility Cloak: Real World Adversarial Attacks on Object Detectors
This addresses security vulnerabilities in object detection systems, with incremental improvements in benchmarking and transferability.
The paper tackles the problem of adversarial attacks on object detectors by training patterns to suppress objectness scores, achieving effectiveness in both digital and physical settings with quantified metrics.
We present a systematic study of adversarial attacks on state-of-the-art object detection frameworks. Using standard detection datasets, we train patterns that suppress the objectness scores produced by a range of commonly used detectors, and ensembles of detectors. Through extensive experiments, we benchmark the effectiveness of adversarially trained patches under both white-box and black-box settings, and quantify transferability of attacks between datasets, object classes, and detector models. Finally, we present a detailed study of physical world attacks using printed posters and wearable clothes, and rigorously quantify the performance of such attacks with different metrics.