RLaGA: A Reinforcement Learning Augmented Genetic Algorithm For Searching Real and Diverse Marker-Based Landing Violations

• URL: http://arxiv.org/abs/2310.07378v2
• Date: Thu, 12 Oct 2023 00:49:09 GMT
• Title: RLaGA: A Reinforcement Learning Augmented Genetic Algorithm For Searching Real and Diverse Marker-Based Landing Violations
• Authors: Linfeng Liang, Yao Deng, Kye Morton, Valtteri Kallinen, Alice James, Avishkar Seth, Endrowednes Kuantama, Subhas Mukhopadhyay, Richard Han, Xi Zheng
• Abstract summary: It's important to fully test auto-landing systems before deploying them in the real-world to ensure safety. This paper proposes RLaGA, a reinforcement learning (RL) augmented search-based testing framework. Our method generates up to 22.19% more violation cases and nearly doubles the diversity of generated violation cases.
• Score: 0.7709288517758135
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Automated landing for Unmanned Aerial Vehicles (UAVs), like multirotor drones, requires intricate software encompassing control algorithms, obstacle avoidance, and machine vision, especially when landing markers assist. Failed landings can lead to significant costs from damaged drones or payloads and the time spent seeking alternative landing solutions. Therefore, it's important to fully test auto-landing systems through simulations before deploying them in the real-world to ensure safety. This paper proposes RLaGA, a reinforcement learning (RL) augmented search-based testing framework, which constructs diverse and real marker-based landing cases that involve safety violations. Specifically, RLaGA introduces a genetic algorithm (GA) to conservatively search for diverse static environment configurations offline and RL to aggressively manipulate dynamic objects' trajectories online to find potential vulnerabilities in the target deployment environment. Quantitative results reveal that our method generates up to 22.19% more violation cases and nearly doubles the diversity of generated violation cases compared to baseline methods. Qualitatively, our method can discover those corner cases which would be missed by state-of-the-art algorithms. We demonstrate that select types of these corner cases can be confirmed via real-world testing with drones in the field.

Related papers

• Variational Autoencoders for exteroceptive perception in reinforcement learning-based collision avoidance [0.0]
  Deep Reinforcement Learning (DRL) has emerged as a promising control framework. Current DRL algorithms require disproportionally large computational resources to find near-optimal policies. This paper presents a comprehensive exploration of our proposed approach in maritime control systems.
  arXiv  Detail & Related papers  (2024-03-31T09:25:28Z)
• Unsupervised Domain Adaptation for Self-Driving from Past Traversal Features [69.47588461101925]
  We propose a method to adapt 3D object detectors to new driving environments. Our approach enhances LiDAR-based detection models using spatial quantized historical features. Experiments on real-world datasets demonstrate significant improvements.
  arXiv  Detail & Related papers  (2023-09-21T15:00:31Z)
• Unsupervised Adaptation from Repeated Traversals for Autonomous Driving [54.59577283226982]
  Self-driving cars must generalize to the end-user's environment to operate reliably. One potential solution is to leverage unlabeled data collected from the end-users' environments. There is no reliable signal in the target domain to supervise the adaptation process. We show that this simple additional assumption is sufficient to obtain a potent signal that allows us to perform iterative self-training of 3D object detectors on the target domain.
  arXiv  Detail & Related papers  (2023-03-27T15:07:55Z)
• You Only Crash Once: Improved Object Detection for Real-Time, Sim-to-Real Hazardous Terrain Detection and Classification for Autonomous Planetary Landings [7.201292864036088]
  A cheap and effective way of detecting hazardous terrain is through the use of visual cameras. Traditional techniques for visual hazardous terrain detection focus on template matching and registration to pre-built hazard maps. We introduce You Only Crash Once (YOCO), a deep learning-based visual hazardous terrain detection and classification technique.
  arXiv  Detail & Related papers  (2023-03-08T21:11:51Z)
• Analyzing Robustness of the Deep Reinforcement Learning Algorithm in Ramp Metering Applications Considering False Data Injection Attack and Defense [0.0]
  Ramp metering is the act of controlling on-going vehicles to the highway mainlines. Deep Q-Learning algorithm uses only loop detectors information as inputs in this study. Model can be applied to almost any ramp metering sites regardless of the road geometries and layouts.
  arXiv  Detail & Related papers  (2023-01-28T00:40:46Z)
• Globally Optimal Boresight Alignment of UAV-LiDAR Systems [0.0]
  misalignments in airborne light detection and ranging (LiDAR) systems can lead to inaccurate 3D point clouds. We introduce a mixed-integer quadratically constrained program (MIQCQP) that can globally solve this misalignment problem. We also propose a nested spatial branch and bound (nsBB) algorithm that improves computational performance.
  arXiv  Detail & Related papers  (2022-02-28T01:48:10Z)
• Rethinking Drone-Based Search and Rescue with Aerial Person Detection [79.76669658740902]
  The visual inspection of aerial drone footage is an integral part of land search and rescue (SAR) operations today. We propose a novel deep learning algorithm to automate this aerial person detection (APD) task. We present the novel Aerial Inspection RetinaNet (AIR) algorithm as the combination of these contributions.
  arXiv  Detail & Related papers  (2021-11-17T21:48:31Z)
• DAE : Discriminatory Auto-Encoder for multivariate time-series anomaly detection in air transportation [68.8204255655161]
  We propose a novel anomaly detection model called Discriminatory Auto-Encoder (DAE) It uses the baseline of a regular LSTM-based auto-encoder but with several decoders, each getting data of a specific flight phase. Results show that the DAE achieves better results in both accuracy and speed of detection.
  arXiv  Detail & Related papers  (2021-09-08T14:07:55Z)
• Exploiting Playbacks in Unsupervised Domain Adaptation for 3D Object Detection [55.12894776039135]
  State-of-the-art 3D object detectors, based on deep learning, have shown promising accuracy but are prone to over-fit to domain idiosyncrasies. We propose a novel learning approach that drastically reduces this gap by fine-tuning the detector on pseudo-labels in the target domain. We show, on five autonomous driving datasets, that fine-tuning the detector on these pseudo-labels substantially reduces the domain gap to new driving environments.
  arXiv  Detail & Related papers  (2021-03-26T01:18:11Z)
• Reinforcement Learning for UAV Autonomous Navigation, Mapping and Target Detection [36.79380276028116]
  We study a joint detection, mapping and navigation problem for a single unmanned aerial vehicle (UAV) equipped with a low complexity radar and flying in an unknown environment. The goal is to optimize its trajectory with the purpose of maximizing the mapping accuracy and to avoid areas where measurements might not be sufficiently informative from the perspective of a target detection.
  arXiv  Detail & Related papers  (2020-05-05T20:39:18Z)
• Physically Realizable Adversarial Examples for LiDAR Object Detection [72.0017682322147]
  We present a method to generate universal 3D adversarial objects to fool LiDAR detectors. In particular, we demonstrate that placing an adversarial object on the rooftop of any target vehicle to hide the vehicle entirely from LiDAR detectors with a success rate of 80%. This is one step closer towards safer self-driving under unseen conditions from limited training data.
  arXiv  Detail & Related papers  (2020-04-01T16:11:04Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.