Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Federated Adversarial Learning for Robust Autonomous Landing Runway Detection

Yi Li, Plamen Angelov, Zhengxin Yu, Alvaro Lopez Pellicer, Neeraj Suri

TL;DR

This paper proposes a federated adversarial learning-based framework to detect landing runways using paired data comprising of clean local data and its adversarial version, and marks the first instance of federated learning work that address the adversarial sample problem in landing runway detection.

Abstract

As the development of deep learning techniques in autonomous landing systems continues to grow, one of the major challenges is trust and security in the face of possible adversarial attacks. In this paper, we propose a federated adversarial learning-based framework to detect landing runways using paired data comprising of clean local data and its adversarial version. Firstly, the local model is pre-trained on a large-scale lane detection dataset. Then, instead of exploiting large instance-adaptive models, we resort to a parameter-efficient fine-tuning method known as scale and shift deep features (SSF), upon the pre-trained model. Secondly, in each SSF layer, distributions of clean local data and its adversarial version are disentangled for accurate statistics estimation. To the best of our knowledge, this marks the first instance of federated learning work that address the adversarial sample problem in landing runway detection. Our experimental evaluations over both synthesis and real images of Landing Approach Runway Detection (LARD) dataset consistently demonstrate good performance of the proposed federated adversarial learning and robust to adversarial attacks.

Federated Adversarial Learning for Robust Autonomous Landing Runway Detection

TL;DR

This paper proposes a federated adversarial learning-based framework to detect landing runways using paired data comprising of clean local data and its adversarial version, and marks the first instance of federated learning work that address the adversarial sample problem in landing runway detection.

Abstract

As the development of deep learning techniques in autonomous landing systems continues to grow, one of the major challenges is trust and security in the face of possible adversarial attacks. In this paper, we propose a federated adversarial learning-based framework to detect landing runways using paired data comprising of clean local data and its adversarial version. Firstly, the local model is pre-trained on a large-scale lane detection dataset. Then, instead of exploiting large instance-adaptive models, we resort to a parameter-efficient fine-tuning method known as scale and shift deep features (SSF), upon the pre-trained model. Secondly, in each SSF layer, distributions of clean local data and its adversarial version are disentangled for accurate statistics estimation. To the best of our knowledge, this marks the first instance of federated learning work that address the adversarial sample problem in landing runway detection. Our experimental evaluations over both synthesis and real images of Landing Approach Runway Detection (LARD) dataset consistently demonstrate good performance of the proposed federated adversarial learning and robust to adversarial attacks.
Paper Structure (24 sections, 6 equations, 3 figures, 4 tables, 1 algorithm)

This paper contains 24 sections, 6 equations, 3 figures, 4 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Works
  3. Deep Learning for ALRD
  4. Adversarial Attacks
  5. Federated Learning
  6. Proposed Approach
  7. Preliminaries
  8. Pre-training
  9. Local Model Fine-tuning
  10. Scale and Shift Deep Features
  11. Adversarial Feature Learning
  12. Central Model Update
  13. Experimental Results
  14. Dataset and Attacks
  15. Competitors and Performance Measure
  16. ...and 9 more sections

Figures (3)

  • Figure 1: Proposed pipeline of the federated learning-based landing runway detection method. The local models are initially pre-trained using lane detection datasets and subsequently fine-tuned with local landing runway detection datasets. The trained Scale and Shift Features (SSF) pools are then aggregated into the final model on the server.
  • Figure 2: Detection error comparison (a) to competitor models (b) against number of clients over LARD.
  • Figure 3: Qualitative landing runway detection results on LARD. From left to right: original images, images attacked by FGSM, results of the central model without adversarial training, results of the central model.