Detection as Regression: Certified Object Detection by Median Smoothing

2020/07/07 by Ping-yeh Chiang, Chiang, Ping-yeh, Michael J. Curry +9 · 2 citations
Computer Science · #Adversarial Robustness in Machine Learning #Computer Vision and Pattern Recognition (cs.CV) #Cryptography and Security (cs.CR) #FOS: Computer and information sciences #Machine Learning (cs.LG)

paper · pdf · doi:10.48550/arxiv.2007.03730

openalex publication_date 2020/07/07 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Despite the vulnerability of object detectors to adversarial attacks, very few defenses are known to date. While adversarial training can improve the empirical robustness of image classifiers, a direct extension to object detection is very expensive. This work is motivated by recent progress on certified classification by randomized smoothing. We start by presenting a reduction from object detection to a regression problem. Then, to enable certified regression, where standard mean smoothing fails, we propose median smoothing, which is of independent interest. We obtain the first model-agnostic, training-free, and certified defense for object detection against ℓ₂-bounded attacks. The code for all experiments in the paper is available at http://github.com/Ping-C/CertifiedObjectDetection .

Citations

• Faster R-CNN: Towards Real-Time Object Detection with Region Proposal\n Networks
• Towards Deep Learning Models Resistant to Adversarial Attacks
• YOLOv3: An Incremental Improvement
• Physical Adversarial Examples for Object Detectors
• On the Effectiveness of Interval Bound Propagation for Training Verifiably Robust Models
• Branch and Bound for Piecewise Linear Neural Network Verification

Cited by

• Fast and Flexible Robustness Certificates for Semantic Segmentation
• Reconcile Certified Robustness and Accuracy for DNN-based Smoothed Majority Vote Classifier
• FS-IQA: Certified Feature Smoothing for Robust Image Quality Assessment

Related

• Certified Adversarial Robustness via Randomized Smoothing
• DetectorGuard: Provably Securing Object Detectors against Localized Patch Hiding Attacks
• FROD: Robust Object Detection for Free
• Towards Bridging the gap between Empirical and Certified Robustness against Adversarial Examples
• Enhancing Object Detection Robustness: Detecting and Restoring Confidence in the Presence of Adversarial Patch Attacks