A Fully-Autonomous Aerial Robot for Search and Rescue Applications in Indoor Environments using Learning-Based Techniques

Sampedro, Carlos; Rodríguez-Ramos, Alejandro; Bavle, Hriday; Carrio, Adrian; Puente, Paloma de la; Campoy, Pascual

doi:10.1007/s10846-018-0898-1

Cited by 157 publications

(84 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…For real-world technologies, such as self-driving cars [10], autonomous drones [14], and search-and-rescue robots [37], the test distribution may be non-stationary, and new observations will often be out-of-distribution (OoD), i.e., not from the training distribution [42]. However, machine learning (ML) models frequently assign wrong labels with high confidence to OoD examples, such as adversarial examples [46,29]-inputs specially crafted by an adversary to cause a target model to misbehave.…”

Section: Introductionmentioning

confidence: 99%

Strike (With) a Pose: Neural Networks Are Easily Fooled by Strange Poses of Familiar Objects

Alcorn

Gong

et al. 2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

198

116

View full text Add to dashboard Cite

Despite excellent performance on stationary test sets, deep neural networks (DNNs) can fail to generalize to out-of-distribution (OoD) inputs, including natural, nonadversarial ones, which are common in real-world settings. In this paper, we present a framework for discovering DNN failures that harnesses 3D renderers and 3D models. That is, we estimate the parameters of a 3D renderer that cause a target DNN to misbehave in response to the rendered image. Using our framework and a self-assembled dataset of 3D objects, we investigate the vulnerability of DNNs to OoD poses of well-known objects in ImageNet. For objects that are readily recognized by DNNs in their canonical poses, DNNs incorrectly classify 97% of their pose space. In addition, DNNs are highly sensitive to slight pose perturbations. Importantly, adversarial poses transfer across models and datasets. We find that 99.9% and 99.4% of the poses misclassified by Inception-v3 also transfer to the AlexNet and ResNet-50 image classifiers trained on the same ImageNet dataset, respectively, and 75.5% transfer to the YOLOv3 object detector trained on MS COCO.

show abstract

Section: Introductionmentioning

confidence: 99%

Strike (With) a Pose: Neural Networks Are Easily Fooled by Strange Poses of Familiar Objects

Alcorn

Gong

et al. 2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

198

116

View full text Add to dashboard Cite

show abstract

“…Machine learning in general and its sub-field deep learning in particular recently reached a remarkable level of supporting practical applications that can effectively utilize these algorithms to enhance the autonomous navigation ability of UAVs as well as optimize UAVs performance in many vital applications, such as object detection and tracking, path planning and navigation, reactive obstacle avoidance, and aggressive maneuvers. [226,[235][236][237][238].…”

Section: F Machine Learningmentioning

confidence: 99%

Design Challenges of Multi-UAV Systems in Cyber-Physical Applications: A Comprehensive Survey and Future Directions

Shakeri

Al-Garadi

Badawy

et al. 2019

IEEE Commun. Surv. Tutorials

204

100

View full text Add to dashboard Cite

Unmanned Aerial Vehicles (UAVs) have recently rapidly grown to facilitate a wide range of innovative applications that can fundamentally change the way cyber-physical systems (CPSs) are designed. CPSs are a modern generation of systems with synergic cooperation between computational and physical potentials that can interact with humans through several new mechanisms. The main advantages of using UAVs in CPS application is their exceptional features, including their mobility, dynamism, effortless deployment, adaptive altitude, agility, adjustability, and effective appraisal of real-world functions anytime and anywhere. Furthermore, from the technology perspective, UAVs are predicted to be a vital element of the development of advanced CPSs. Therefore, in this survey, we aim to pinpoint the most fundamental and important design challenges of multi-UAV systems for CPS applications. We highlight key and versatile aspects that span the coverage and tracking of targets and infrastructure objects, energy-efficient navigation, and image analysis using machine learning for fine-grained CPS applications. Key prototypes and testbeds are also investigated to show how these practical technologies can facilitate CPS applications. We present and propose state-of-the-art algorithms to address design challenges with both quantitative and qualitative methods and map these challenges with important CPS applications to draw insightful conclusions on the challenges of each application. Finally, we summarize potential new directions and ideas that could shape future research in these areas. M 2 respectively. However, as shown in Fig.1, both technologies provide two extremes for surveillance with respect to accuracy versus intrusive deployment and maintenance. SRS provides slow large-scale monitoring, while CSNs require widespread deployment and constant human intervention for setup and maintenance, which may not be feasible particularly in harsh environments. On the one hand, SRS cannot be used for finegrain monitoring, such as monitoring of structural objects, with high accuracy, which limits its usage in many applications that require close monitoring and context detection [5]. On the other hand, traditional sensor networks (SNs) are spreading in countless number of Internet of Things (IoT) applications to provide automatic monitoring of objects, humans, and/or infrastructure. SNs, however, require permanent deployment of large magnitude of localized sensors, intrusively to the environment in many cases, and also require constant human intervention to manage and maintain. The proliferation of lowcost UAVs [6] is emerging to provide innovative ways of low altitude sensing with zero-deployment (i.e., no fixed deployment requirements) that can fundamentally change the way such applications are designed.Similar to surveillance applications, UAV technologies have progressed significantly in several applications in the past decade, making commercial and civil applications feasible and compelling. As it occurs with many disruptive technolo...

show abstract

“…Whereas such ground obstacles would be detected and avoided using the 2D laser sensor on board the UAV. Obstacle detection and avoidance using the 2D laser can be found in our previous work [24].…”

Section: Limitationsmentioning

confidence: 99%

Fast and Robust Flight Altitude Estimation of Multirotor UAVs in Dynamic Unstructured Environments Using 3D Point Cloud Sensors

et al. 2018

Self Cite

View full text Add to dashboard Cite

This paper presents a fast and robust approach for estimating the flight altitude of multirotor Unmanned Aerial Vehicles (UAVs) using 3D point cloud sensors in cluttered, unstructured, and dynamic indoor environments. The objective is to present a flight altitude estimation algorithm, replacing the conventional sensors such as laser altimeters, barometers, or accelerometers, which have several limitations when used individually. Our proposed algorithm includes two stages: in the first stage, a fast clustering of the measured 3D point cloud data is performed, along with the segmentation of the clustered data into horizontal planes. In the second stage, these segmented horizontal planes are mapped based on the vertical distance with respect to the point cloud sensor frame of reference, in order to provide a robust flight altitude estimation even in presence of several static as well as dynamic ground obstacles. We validate our approach using the IROS 2011 Kinect dataset available in the literature, estimating the altitude of the RGB-D camera using the provided 3D point clouds. We further validate our approach using a point cloud sensor on board a UAV, by means of several autonomous real flights, closing its altitude control loop using the flight altitude estimated by our proposed method, in presence of several different static as well as dynamic ground obstacles. In addition, the implementation of our approach has been integrated in our open-source software framework for aerial robotics called Aerostack.

show abstract

A Fully-Autonomous Aerial Robot for Search and Rescue Applications in Indoor Environments using Learning-Based Techniques

Cited by 157 publications

References 39 publications

Strike (With) a Pose: Neural Networks Are Easily Fooled by Strange Poses of Familiar Objects

Strike (With) a Pose: Neural Networks Are Easily Fooled by Strange Poses of Familiar Objects

Design Challenges of Multi-UAV Systems in Cyber-Physical Applications: A Comprehensive Survey and Future Directions

Fast and Robust Flight Altitude Estimation of Multirotor UAVs in Dynamic Unstructured Environments Using 3D Point Cloud Sensors

Contact Info

Product

Resources

About