DroNet: Learning to Fly by Driving

By Antonio Loquercio, Ana I. Maqueda, Carlos R. del-Blanco, Davide Scaramuzza. 2018

463 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Automotive Engineering, Computer Vision and Pattern Recognition, Robotics & Control

How researchers see this paper

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

In Artificial Intelligence

In Automotive Engineering

In Computer Vision and Pattern Recognition

In Robotics & Control

Citations in the corpus

A General Framework for Uncertainty Estimation in Deep Learning (2019, methodology): “[31], since it was shown to allow closedloop control on several unmanned aerial platforms [32], [33].”
Beauty and the Beast: Optimal Methods Meet Learning for Drone Racing (2018, methodology): “First, the input image is processed by a Convolutional Neural Network (CNN), based on the shallow DroNet architecture [15].”
Deep Drone Racing: Learning Agile Flight in Dynamic Environments (2018, background): “Known output representations range from predicting discrete navigation commands [15, 17]— which enables high robustness but leads to low agility—to direct control [14]—which can lead to highly agile control, but suffers from high sample complexity.”
A 64-mW DNN-Based Visual Navigation Engine for Autonomous Nano-Drones (2018, methodology): “[2] for standard-sized UAVs, to the computational requirements of a nano-sized UAV, such as fixed-point computation; 3) we deployed DroNet on the PULP-Shield, an ultralow-power visual navigation module featuring the GAP8 SoC, an ultralow-power camera and off-chip Flash/DRAM memory; the shield is des”
BADGR: An Autonomous Self-Supervised Learning-Based Navigation System (2020, background): “Prior works have investigated learning navigation policies directly from real-world experience, but typically require a person [30, 33, 34] or SLAM algorithm [35] to gather the data, assume access to the ground-truth robot state [36], learn using low-bandwidth sensors [37], or only perform collision”
Learning to Drive in a Day (2018, background): “Variants of this approach have been demonstrated on self-driving cars [18], [19] and drones [20], [21], [22].”
Deep Local Trajectory Replanning and Control for Robot Navigation (2019, methodology): “[8, 9, 10, 11, 12] and hierarchical planning [13] to improve reactive robot control.”
GONet: A Semi-Supervised Deep Learning Approach For Traversability Estimation (2018, background): “It is worth noting that prior work involving learned motion control policies for robots have started to reason about the traversability of environments along with a robot’s next actions [42], [43], [23], [21].”
Combining Optimal Control and Learning for Visual Navigation in Novel Environments (2019, background): “A number of papers seek to combine the best of learning with optimal control for high-speed navigation [27, 28, 29, 30, 31].”
ViNT: A Foundation Model for Visual Navigation (2023, background): “Instead, our paper follows the paradigm of learning navigation behavior from data collected across multiple different real-world robotic systems [17–19], while focusing on training a foundation model that can be adapted for a variety of downstream tasks in zero shot or with small amounts of data.”
Learning high-speed flight in the wild (2021, background): “Because the number of samples required to train general navigation policies is very high, existing approaches impose constraints on the quadrotor’s motion model, for example, by constraining the platform to planar motion (24, 26, 27) and/or discrete actions (25), at the cost of reduced maneuverabili”
Recent Advances in Robot Learning from Demonstration (2020, background): “Furthermore, recent advances in deep learning have fueled the development of end-to-end LfD methods for aerial vehicles (175, 176).”