Toward Low-Flying Autonomous MAV Trail Navigation using Deep Neural Networks for Environmental Awareness

被引:0
|
作者
Smolyanskiy, Nikolai [1 ]
Kamenev, Alexey [1 ]
Smith, Jeffrey [1 ]
Birchfield, Stan [1 ]
机构
[1] NVIDIA Corp, Redmond, WA 98052 USA
关键词
VISION;
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
We present a micro aerial vehicle (MAV) system, built with inexpensive off-the-shelf hardware, for autonomously following trails in unstructured, outdoor environments such as forests. The system introduces a deep neural network (DNN) called TrailNet for estimating the view orientation and lateral offset of the MAV with respect to the trail center. The DNN-based controller achieves stable flight without oscillations by avoiding overconfident behavior through a loss function that includes both label smoothing and entropy reward. In addition to the TrailNet DNN, the system also utilizes vision modules for environmental awareness, including another DNN for object detection and a visual odometry component for estimating depth for the purpose of low-level obstacle detection. All vision systems run in real time on board the MAV via a Jetson TX1. We provide details on the hardware and software used, as well as implementation details. We present experiments showing the ability of our system to navigate forest trails more robustly than previous techniques, including autonomous flights of 1 km.
引用
收藏
页码:4241 / 4247
页数:7
相关论文
共 50 条
  • [1] Autonomous UAV Trail Navigation with Obstacle Avoidance Using Deep Neural Networks
    Seungho Back
    Gangik Cho
    Jinwoo Oh
    Xuan-Toa Tran
    Hyondong Oh
    [J]. Journal of Intelligent & Robotic Systems, 2020, 100 : 1195 - 1211
  • [2] Autonomous UAV Trail Navigation with Obstacle Avoidance Using Deep Neural Networks
    Back, Seungho
    Cho, Gangik
    Oh, Jinwoo
    Tran, Xuan-Toa
    Oh, Hyondong
    [J]. JOURNAL OF INTELLIGENT & ROBOTIC SYSTEMS, 2020, 100 (3-4) : 1195 - 1211
  • [3] Autonomous navigation using neural networks
    Deming, JR
    de Oliveira, MAA
    [J]. PROCEEDINGS OF THE 2004 INTERNATIONAL CONFERENCE ON MACHINE LEARNING AND APPLICATIONS (ICMLA'04), 2004, : 235 - 241
  • [4] Waypoint reduction to improve autonomous navigation using deep neural networks and path planners
    Gayathri, R.
    Uma, V
    O'Brien, Bettina
    [J]. SADHANA-ACADEMY PROCEEDINGS IN ENGINEERING SCIENCES, 2024, 49 (02):
  • [5] Neural networks filter for hybrid navigation of formation flying spacecrafts in deep space
    Li Hui
    Zhang Qinyu
    Zhang Naitong
    [J]. SECOND INTERNATIONAL CONFERENCE ON SPACE INFORMATION TECHNOLOGY, PTS 1-3, 2007, 6795
  • [6] Monocular vision with deep neural networks for autonomous mobile robots navigation
    Sleaman, Walead Kaled
    Hameed, Alaa Ali
    Jamil, Akhtar
    [J]. OPTIK, 2023, 272
  • [7] Navigation of autonomous robots using fuzzy-neural networks
    Markusek, J
    Vitko, A
    Jurisica, L
    [J]. PROCEEDINGS OF THE 2ND INTERNATIONAL CONFERENCE ON CLIMBING AND WALKING ROBOTS, CLAWAR 99, 1999, : 123 - 131
  • [8] Autonomous navigation of formation flying spacecrafts in deep space exploration and communication by hybrid navigation utilizing neural network filter
    Li, H.
    Zhang, Q. Y.
    Zhang, N. T.
    [J]. ACTA ASTRONAUTICA, 2009, 65 (7-8) : 1028 - 1031
  • [9] Moving toward autonomous manufacturing by accelerating hydrodynamic fabrication of microstructures using deep neural networks
    Clinkinbeard, Nicholus R.
    Hashemi, Nicole N.
    [J]. MICRO AND NANO ENGINEERING, 2024, 24
  • [10] Sensor fusion for the navigation of Autonomous Guided Vehicle using neural networks
    Cao, J
    Hall, E
    [J]. INTELLIGENT ROBOTS AND COMPUTER VISION XVII: ALGORITHMS, TECHNIQUES, AND ACTIVE VISION, 1998, 3522 : 286 - 294