219 Hits in 4.9 sec

Flying Free: A Research Overview of Deep Learning in Drone Navigation Autonomy

Thomas Lee, Susan Mckeever, Jane Courtney
2021 Drones  
in order to create a clear definition of autonomy when applied to drones.  ...  Research activity in autonomous drone navigation has increased rapidly in the past five years, and drones are moving fast towards the ultimate goal of near-complete autonomy.  ...  A Convolutional Neural Network Feature Detection Approach to Autonomous Quadrotor Indoor Navigation.  ... 
doi:10.3390/drones5020052 fatcat:jqel25c655ajrkd2kzyyucy6ku

A Review on IoT Deep Learning UAV Systems for Autonomous Obstacle Detection and Collision Avoidance

Paula Fraga-Lamas, Lucía Ramos, Víctor Mondéjar-Guerra, Tiago M. Fernández-Caramés
2019 Remote Sensing  
autonomous and robust real-time obstacle detection and avoidance systems.  ...  In this regard, Deep Learning (DL) techniques have arisen as a promising alternative for improving real-time obstacle detection and collision avoidance for highly autonomous UAVs.  ...  On the one hand, a Convolutional Neural Network (CNN) is a type of hierarchical architecture for feature extraction that allows for generating a set of high-dimensional feature maps from the raw sensor  ... 
doi:10.3390/rs11182144 fatcat:54xs26xnvzf7rfa5b64tuzkz44

Deep Reinforcement Learning for Autonomous Map-less Navigation of a Flying Robot

Oualid Doukhi, Deok Jin Lee
2022 IEEE Access  
It transforms the fused sensory data to a velocity control input for the robot through an end-to-end Convolutional Neural Network (CNN).  ...  This paper proposes a novel approach for enabling a Micro Aerial Vehicle (MAV) system equipped with a laser rangefinder and depth sensor to autonomously navigate and explore an unknown indoor or outdoor  ...  CONCLUSION This paper presents a novel approach for integrated autonomous navigation and obstacle avoidance for a MAV quadrotor using a modular Deep Q-Network architecture.  ... 
doi:10.1109/access.2022.3162702 fatcat:6s2vyhbnxzbqhaupfbo2xod6hq

Deep Learning and Reinforcement Learning for Autonomous Unmanned Aerial Systems: Roadmap for Theory to Deployment [article]

Jithin Jagannath, Anu Jagannath, Sean Furman, Tyler Gwin
2020 arXiv   pre-print
Then we discuss how reinforcement learning is explored for using this information to provide autonomous control and navigation for UAS.  ...  A key area of focus that will be essential to enable autonomy to UAS is computer vision.  ...  Convolutional Neural Networks Convolutional networks or convolutional neural networks (CNNs) are a specialized type of feedforward neural network that performs convolution operation in at least one of  ... 
arXiv:2009.03349v2 fatcat:5ylreoukrfcrtorzzp44mntjum


2021 2021 18th International Conference on Electrical Engineering, Computing Science and Automatic Control (CCE)  
inequalities A Generalized Lagrange Multiplier Method for Support Vector Regression Convolutional neural networks Semantic Segmentation of Mammograms Using Pre-Trained Deep Neural Networks Cu12Sb4S13  ...  screw Autonomous navigation of a mobile robot using a network of Hindmarsh-Rose (HR) neurons A Novel Method to Analyze Input-Output Controllability A systematic method for backstepping via linear matrix  ...  Spine surgery Interpedicular screw placement image guided navigation surgery simulator.  ... 
doi:10.1109/cce53527.2021.9633101 fatcat:7ffdhuyqevhmpawcjajs2rgniq

CNN-Based Vision Model for Obstacle Avoidance of Mobile Robot

Canglong Liu, Bin Zheng, Chunyang Wang, Yongting Zhao, Shun Fu, Haochen Li, Bing Xu, Yinong Chen
2017 MATEC Web of Conferences  
We present an end-to-end learning model based Convolutional Neural Network (CNN), which takes the raw image obtained from camera as only input.  ...  Our neural network was trained under caffe framework and specific instructions are executed by the Robot Operating System (ROS).  ...  They trained a end-to-end learning model based convolutional neural network. The image obtained from a camera as input and output the command for control the robot arm.  ... 
doi:10.1051/matecconf/201713900007 fatcat:3nd6cgsjgjbhrlg5kzrncwok6q

QuadNet: A Hybrid Framework for Quadrotor Dead Reckoning

Artur Shurin, Itzik Klein
2022 Sensors  
As a hybrid approach, QuadNet uses both neural networks and model-based equations during its operation. QuadNet requires only the inertial sensor readings to provide the position vector.  ...  Commonly, quadrotors use an inertial navigation system combined with a global navigation satellite systems receiver for outdoor applications and a camera for indoor/outdoor applications.  ...  As a hybrid approach, QuadNet uses both neural network (NN) and model-based equations.  ... 
doi:10.3390/s22041426 pmid:35214328 pmcid:PMC8878889 fatcat:ggrcd2xainaxbb3ru64w4xphie

Obstacle Avoidance through Deep Networks based Intermediate Perception [article]

Shichao Yang, Sandeep Konam, Chen Ma, Stephanie Rosenthal, Manuela Veloso, Sebastian Scherer
2017 arXiv   pre-print
The Convolutional Neural Network perception is divided into two stages: first, predict depth map and surface normal from RGB images, which are two important geometric properties related to 3D obstacle  ...  Our model generalizes well to other public indoor datasets and is also demonstrated for robot flights in simulation and experiments.  ...  To navigate autonomously, the vehicles need to detect and avoid 3D obstacles in real time.  ... 
arXiv:1704.08759v1 fatcat:a26dpm3rujeqpcfdfvdb2enwxa

CAD2RL: Real Single-Image Flight Without a Single Real Image

Fereshteh Sadeghi, Sergey Levine
2017 Robotics: Science and Systems XIII  
ACKNOWLEDGMENT The authors would like to thank Larry Zitnick for helpful discussions and insightful remarks.  ...  As the result, the network learns geometric features and can robustly detect open spaces.  ...  We randomize textures, lighting and furniture placement to create a visually diverse set of scenes. Fig. 3 . We use a fully convolutional neural network to learn the Q-function.  ... 
doi:10.15607/rss.2017.xiii.034 dblp:conf/rss/SadeghiL17 fatcat:meuops6hdzb4pbfluhw2hgv5gu

Beauty and the Beast: Optimal Methods Meet Learning for Drone Racing [article]

Elia Kaufmann, Mathias Gehrig, Philipp Foehn, René Ranftl, Alexey Dosovitskiy, Vladlen Koltun, Davide Scaramuzza
2019 arXiv   pre-print
At test time, a convolutional network predicts the poses of the closest gates along with their uncertainty.  ...  The presented approach was used to win the IROS 2018 Autonomous Drone Race Competition, outracing the second-placing team by a factor of two.  ...  First, the input image is processed by a Convolutional Neural Network (CNN), based on the shallow DroNet architecture [15] .  ... 
arXiv:1810.06224v4 fatcat:4flvlfkx2jfmjmymdp3f5e62vi

Convolutional Neural Network-Based Robot Navigation Using Uncalibrated Spherical Images

Lingyan Ran, Yanning Zhang, Qilin Zhang, Tao Yang
2017 Sensors  
The classification is implemented as an end-to-end Convolutional Neural Network (CNN), trained on our proposed Spherical-Navi image dataset, whose category labels can be efficiently collected.  ...  Another category of human vision inspired approaches, i.e., the robot-oriented heading-field road detection and trajectory planning methods, address the navigation problem directly with visual paths detection  ...  In this section, a convolutional neural network based robot navigation framework is formulated to accurately estimate robot heading direction using raw spherical images.  ... 
doi:10.3390/s17061341 pmid:28604624 pmcid:PMC5492478 fatcat:owcdps6okrfb5olmvzz5b5jkcq

CAD2RL: Real Single-Image Flight without a Single Real Image [article]

Fereshteh Sadeghi, Sergey Levine
2017 arXiv   pre-print
Our learned collision avoidance policy is represented by a deep convolutional neural network that directly processes raw monocular images and outputs velocity commands.  ...  We evaluate our method by flying a real quadrotor through indoor environments, and further evaluate the design choices in our simulator through a series of ablation studies on depth prediction.  ...  ACKNOWLEDGMENT The authors would like to thank Larry Zitnick for helpful discussions and insightful remarks.  ... 
arXiv:1611.04201v4 fatcat:7w3vienpabgu7ejsfsq2qkqesm

A 64mW DNN-based Visual Navigation Engine for Autonomous Nano-Drones

Daniele Palossi, Antonio Loquercio, Francesco Conti, Francesco Conti, Eric Flamand, Eric Flamand, Davide Scaramuzza, Luca Benini, Luca Benini
2019 IEEE Internet of Things Journal  
Visual navigation based on AI approaches, such as deep neural networks (DNNs) are becoming pervasive for standard-size drones, but are considered out of reach for nanodrones with size of a few cm^2.  ...  As part of our general methodology we discuss the software mapping techniques that enable the state-of-the-art deep convolutional neural network presented in [1] to be fully executed on-board within a  ...  the CNN-based approach for autonomous navigation proposed in Loquercio et al  ... 
doi:10.1109/jiot.2019.2917066 fatcat:ogqpf3qzg5hc5hph6cgdccivxu

Application Challenges from a Bird's-Eye View [chapter]

Davide Scaramuzza
2017 Computer Vision in Vehicle Technology  
, lane departure warning, traffic sign recognition), autonomous driving and robot navigation (with visual simultaneous localization and mapping) or unmanned aerial vehicles (obstacle avoidance, landscape  ...  It also serves as a reference for researchers of current developments and challenges in areas of the application of computer vision, involving vehicles such as advanced driver assistance (pedestrian detection  ...  6 .4 Autonomous recovery after throwing the quadrotor by hand: (a) the quadrotor detects free fall and (b) starts to control its attitude to be horizontal.  ... 
doi:10.1002/9781118868065.ch6 fatcat:fsljmro5izhh5mp4z3hcrgstby

Deep learning for vision-based micro aerial vehicle autonomous landing

Leijian Yu, Cai Luo, Xingrui Yu, Xiangyuan Jiang, Erfu Yang, Chunbo Luo, Peng Ren
2018 International Journal of Micro Air Vehicles  
To overcome these limitations, we propose an end-to-end landmark detection system based on a deep convolutional neural network, which not only easily scales up to a larger number of various landmarks but  ...  Furthermore, we propose a separative implementation strategy which conducts convolutional neural network training and detection on different hardware platforms separately, i.e. a graphics processing unit  ...  Training a convolutional neural network for landmark detection Inspired by the Yolo model 17 and SqueezeNet 22 modeling methodologies, we develop a convolutional neural network that performs end-to-end  ... 
doi:10.1177/1756829318757470 fatcat:asq6ksh52bdirign3ahalewz3y
« Previous Showing results 1 — 15 out of 219 results