Toward low-flying autonomous MAV trail navigation using deep neural networks for environmental awareness

By Nikolai Smolyanskiy, Alexey Kamenev, Jeffrey Smith, Stan Birchfield. 2017

240 citations recorded by the source corpus.

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.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

In Computer Vision and Pattern Recognition

Citations in the corpus

A 64-mW DNN-Based Visual Navigation Engine for Autonomous Nano-Drones (2018, background): “Recent results have shown that much lighter algorithms, based on convolutional neural networks (CNNs), are sufﬁcient for enabling basic reactive navigation of small drones, even without a map of the environment [2], [17], [18], [19], [20].”
DroNet: Learning to Fly by Driving (2018, background): “other CNN-based controllers [9], [13], [25].”
SPINN: synergistic progressive inference of neural networks over device and cloud (2020, background): “Furthermore, emerging user-facing [80] and missioncritical [37, 42, 66] CNN applications require low-latency processing to ensure high quality of experience (QoE) [4] and safety [11].”
Sim4CV: A Photo-Realistic Simulator for Computer Vision Applications (2017, unknown): “− This is also the setting used in several related work, including [50] and probably [4] (two rotational oﬀsets only are used in both papers, but the exact angles are not given in the latter).”
Teaching UAVs to Race Using UE4Sim (2017, methodology): “For the particular case of physics-driven vehicles, SL can be advantageous when high dimensional feedback can be recorded and it has proven to have relative success in the last several years in the field of autonomous driving [3,4,40].”
Feature Distillation: DNN-Oriented JPEG Compression Against Adversarial Examples (2018, background): “…image recognition, adversarially manipulating the perceptual systems of autonomous vehicles by physically captured adversarial images, i.e. via camera or sensor [17, 22], can lead to the misreading on road signs, thus causing potential disas-trous consequences in DNN-based cyber-physical systems.”
Teaching UAVs to Race: End-to-End Regression of Agile Controls in Simulation (2017, methodology): “Navigation has traditionally been approached by either employing supervised learning (SL) methods [3,4,17,29,35,38,42] or reinforcement learning (RL) methods [21, 24, 25, 32, 33, 43].”