FROD: Robust Object Detection for Free

TL;DR

This paper introduces FROD, a method to improve the robustness of object detection models against adversarial attacks by modifying classification backbones and adding lightweight components, achieving robustness without extra computational cost.

Contribution

The paper proposes a novel approach to enhance object detection robustness by modifying classification backbones and introducing lightweight training components, bridging the gap with classification robustness.

Findings

Enhanced robustness on MS-COCO and Pascal VOC datasets.

Effective modifications do not increase computational overhead.

Lightweight components further improve robustness.

Abstract

Object detection is a vital task in computer vision and has become an integral component of numerous critical systems. However, state-of-the-art object detectors, similar to their classification counterparts, are susceptible to small adversarial perturbations that can significantly alter their normal behavior. Unlike classification, the robustness of object detectors has not been thoroughly explored. In this work, we take the initial step towards bridging the gap between the robustness of classification and object detection by leveraging adversarially trained classification models. Merely utilizing adversarially trained models as backbones for object detection does not result in robustness. We propose effective modifications to the classification-based backbone to instill robustness in object detection without incurring any computational overhead. To further enhance the robustness achieved by the proposed modified backbone, we introduce two lightweight components: imitation loss and delayed adversarial training. Extensive experiments on the MS-COCO and Pascal VOC datasets are conducted to demonstrate the effectiveness of our proposed approach.