Learning monocular reactive UAV control in cluttered natural environments

By Stéphane Ross, Narek Melik-Barkhudarov, Kumar Shaurya Shankar, Andreas Wendel, Debadeepta Dey, J. Bagnell, Martial Hebert. 2012

452 citations recorded by the source corpus.

Autonomous navigation for large Unmanned Aerial Vehicles (UAVs) is fairly straight-forward, as expensive sensors and monitoring devices can be employed. In contrast, obstacle avoidance remains a challenging task for Micro Aerial Vehicles (MAVs) which operate at low altitude in cluttered environments. Unlike large vehicles, MAVs can only carry very light sensors, such as cameras, making autonomous navigation through obstacles much more challenging. In this paper, we describe a system that navigates a small quadrotor helicopter autonomously at low altitude through natural forest environments. Using only a single cheap camera to perceive the environment, we are able to maintain a constant velocity of up to 1.5m/s. Given a small set of human pilot demonstrations, we use recent state-of-the-art imitation learning techniques to train a controller that can avoid trees by adapting the MAVs heading. We demonstrate the performance of our system in a more controlled environment indoors, and in real natural forest environments outdoors.

Subfields: Aerospace 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 Aerospace Engineering

In Computer Vision and Pattern Recognition

In Robotics & Control

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]–[20].”
BADGR: An Autonomous Self-Supervised Learning-Based Navigation System (2020, background): “Meth-ods based on imitation learning have been demonstrated on real-world robots [16], [17], but again depend on humans for expert demonstrations, which does not constitute a continuously self-improving system.”
End-to-End Driving Via Conditional Imitation Learning (2017, background): “Imitation learning has been applied to a variety of tasks, including articulated motion [2], [28], [11], autonomous flight [1], [13], [30], modeling navigational behavior [37], [38], off-road driving [22], [32], and road following [4], [6], [27], [36].”
Agile Autonomous Driving using End-to-End Deep Imitation Learning (2017, methodology): “Therefore, the frame-byframe labeling approach [26], for example, can lead to a very counter-intuitive, inefficient data collection process, because the required dynamics information is lost in a single image frame.”
(CAD)$^2$RL: Real Single-Image Flight without a Single Real Image (2016, background): “Conventionally, learning-based approaches to autonomous flight have relied on learning from demonstration [2, 1, 30, 34].”
Learning to Act by Predicting the Future (2016, methodology): “While the use of SL is natural in imitation learning (LeCun et al., 2005; Ross et al., 2013) or in conjunction with model-based RL (Levine & Koltun, 2013), the formulation of sensorimotor learning from raw experience as supervised learning is rare (Levine et al., 2016).”
Visual Imitation Made Easy (2020, methodology): “It has been successfully applied in solving a wide range of tasks including playing games [28], self-driving [23], and navigating drones through cluttered environments [22].”
IRIS: Implicit Reinforcement without Interaction at Scale for Learning Control from Offline Robot Manipulation Data (2019, methodology): “The loss function for the RNN is a simple Behavioral Cloning loss L θ ( a t : t + T , s t : t + T .”
Multiple Interactions Made Easy (MIME): Large Scale Demonstrations Data for Imitation (2018, background): “This has enabled autonomous helicopter aerobatics [35], table-tennis playing [36] and drone ﬂying [37].”
Deep Imitation Learning for Complex Manipulation Tasks from Virtual Reality Teleoperation (2017, background): “This approach has proven successful in domains where collecting such demonstrations is natural, such as simulated environments [24], [35], driving [4], [5], [6], and drones [36].”
An Algorithmic Perspective on Imitation Learning (2018, methodology): “Right: Learning autonomous UAV flight [Ross et al., 2013].”
PLATO: Policy learning using adaptive trajectory optimization (2016, neutral): “expert [15] or a computational expert, such as Monte Carlo tree search [16].”
Query-Efficient Imitation Learning for End-to-End Simulated Driving (2017, methodology): “First, as noted in (Ross et al. 2013), a human operator cannot drive well without actual feedback, which is the case of DAgger as the primary policy drives most of the time.”
Query-Efficient Imitation Learning for End-to-End Autonomous Driving (2016, methodology): “First, as noted in [17], a human operator cannot drive well without actual feedback, which is the case of DAgger as the primary policy drives most of the time.”
Comparing human-centric and robot-centric sampling for robot deep learning from demonstrations (2016, methodology): “However in [20], it was recommended to display slowed down videos of the robot’s execution to the human, which prohibits stochastic mixing.”
Imitation Learning as f-Divergence Minimization (2019, background): “Although shown to be successful in various applications [1, 18, 19], there are domains where it’s impractical to have on-policy expert labels [20, 21].”
Spatial Action Maps for Mobile Manipulation (2020, background): “Other work has considered the impact of different state representations: from front-facing camera images [4, 22, 24, 26] to top-down bird’s-eye views (BEV) of the scene generated with Inverse Perspective Mapping (IPM) [5, 7, 11, 29].”
Deep Neural Network for Real-Time Autonomous Indoor Navigation (2015, unknown): “Finally, section 6 oﬀers concluding remarks.”