Making an Invisibility Cloak: Real World Adversarial Attacks on Object Detectors
arXiv:1910.14667
Abstract
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.
ECCV 2020
References in corpus (6)
- Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
- Robust Physical-World Attacks on Deep Learning Models
- AdvHat: Real-world adversarial attack on ArcFace Face ID system
- DARTS: Deceiving Autonomous Cars with Toxic Signs
- Robust Adversarial Perturbation on Deep Proposal-based Models
- Exploring the Vulnerability of Single Shot Module in Object Detectors via Imperceptible Background Patches