Introduction

Deep neural network (DNN) based object detection has become an integral part of numerous cyber-physical systems, perceiving physical environments and responding proactively to real-time events. Recent studies reveal that well-trained multi-task learners like DNN-based object detectors perform poorly in the presence of deception. This project presents FUSE, a deception-resilient detection fusion approach with diversity-enhanced fusion teaming mechanisms, including diversity-enhanced joint training algorithms, for producing high diversity fusion detectors. We develop a three-tier detection fusion framework and a graph partition algorithm to construct fusion-verified detection outputs through three mutually-reinforcing components: objectness fusion, bounding box fusion, and classification fusion. In our paper to be presented at SIGKDD 2021, we provide a formal analysis of robustness enhancement by FUSE-protected systems and extensive experimental results on eleven detectors from three families of detection algorithms on two benchmark datasets. We show that FUSE guarantees strong robustness in mitigating the state-of-the-art deception attacks, including adversarial patches - a form of physical attacks using confined visual distortion.

This repository contains the source code optimized on top of the following paper:

Installation and Dependencies

This project runs on Python 3.6. It is lightweight and only requires the library itertools that is pre-installed on most Python-enabled environment.

Instruction

  1. Prepare a list of candidate objects detected by member models. The example below has three members: (1) Faster RCNN contributing one candidate, (2) SSD512 contributing two candidates, and (3) YOLOv3 (Darknet53) contributing four candidates. Note that you should create the Object instances as shown below. ```python from tools.graph import Object

candidates = [ Object(model=’FRCNN’, class_name=’train’, confidence=1.00, xmin=125, ymin=179, xmax=215, ymax=313), Object(model=’SSD512’, class_name=’train’, confidence=1.00, xmin=138, ymin=197, xmax=205, ymax=300), Object(model=’SSD512’, class_name=’pottedplant’, confidence=0.24, xmin=202, ymin=238, xmax=225, ymax=294), Object(model=’YOLOv3 (Darknet53)’, class_name=’train’, confidence=0.98, xmin=135, ymin=193, xmax=211, ymax=299), Object(model=’YOLOv3 (Darknet53)’, class_name=’person’, confidence=0.07, xmin=144, ymin=234, xmax=155, ymax=248), Object(model=’YOLOv3 (Darknet53)’, class_name=’person’, confidence=0.01, xmin=187, ymin=234, xmax=195, ymax=248), Object(model=’YOLOv3 (Darknet53)’, class_name=’car’, confidence=0.01, xmin=229, ymin=228, xmax=239, ymax=242) ]


2. Run `fuse` to obtain the fusion results which contain a list of `Object` instances
```python
from core.defense import fuse

fusion_results = fuse(candidates)

Note: You may need to edit the tools/constants.py file to specify the total number of members according to your system configuration and the IOU threshold for hyperparameter tuning.

Status

The code is provided as is, without warranty or support. If you use our code, please cite:

@inproceedings{chow2021robust,
  title={Robust Object Detection Fusion Against Deception},
  author={Chow, Ka-Ho and Liu, Ling},
  booktitle={ACM SIGKDD International Conference on Knowledge Discovery and Data Mining},
  year={2021},
  organization={ACM}
}

We are continuing the development and there is ongoing work in our lab regarding robust object detection. If you would like to contribute to this project, please contact Ka-Ho Chow.

@inproceedings{chow2020adversarial,
  title={Adversarial Objectness Gradient Attacks in Real-time Object Detection Systems},
  author={Chow, Ka-Ho and Liu, Ling and Loper, Margaret and Bae, Juhyun and Emre Gursoy, Mehmet and Truex, Stacey and Wei, Wenqi and Wu, Yanzhao},
  booktitle={IEEE International Conference on Trust, Privacy and Security in Intelligent Systems, and Applications},
  pages={263--272},
  year={2020},
  organization={IEEE}
}

@inproceedings{chow2020understanding,
  title={Understanding Object Detection Through an Adversarial Lens},
  author={Chow, Ka-Ho and Liu, Ling and Gursoy, Mehmet Emre and Truex, Stacey and Wei, Wenqi and Wu, Yanzhao},
  booktitle={European Symposium on Research in Computer Security},
  pages={460--481},
  year={2020},
  organization={Springer}
}

Acknowledgement

This project is developed based on the following repository: