High speed obstacle avoidance using monocular vision and reinforcement learning

By Jeff Michels, Ashutosh Saxena, Andrew Y. Ng. 2005

473 citations recorded by the source corpus.

Subfields: Aerospace Engineering, 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 Aerospace Engineering

In Automotive Engineering

In Computer Vision and Pattern Recognition

In Robotics & Control

Citations in the corpus

Reinforcement learning in robotics: A survey (2013, methodology): “It has been used to learn control parameters for a RC car (Michels et al., 2005) and an autonomous blimp (Ko et al.”
Depth Map Prediction from a Single Image using a Multi-Scale Deep Network (2014, background): “In turn, such relations help provide richer representations of objects and their environment, often leading to improvements in existing recognition tasks [18], as well as enabling many further applications such as 3D modeling [16, 6], physics and support models [18], robotics [4, 14], and potentiall”
Driving Policy Transfer via Modularity and Abstraction (2018, methodology): “for lane following using synthetic road images and then deployed it on a physical vehicle. This early work is inspiring, but is restricted to rudimentary lane following. More recently, Michels et al. [31] trained an obstacle avoidance policy in simulation and transferred it onto a small robotic vehi”
Self-Supervised Deep Reinforcement Learning with Generalized Computation Graphs for Robot Navigation (2017, methodology): “While these methods have been used to learn robot navigation policies, they often require simulation experience [19], [20].”
BADGR: An Autonomous Self-Supervised Learning-Based Navigation System (2020, methodology): “nput sensory data with semantically meaningful labels, such as which pixels in an image correspond to traversable or bumpy terrain. However, these methods typically depend on existing SLAM approaches [4, 5, 23, 9, 24] or humans [6, 7] in order to provide the semantic labels, which consequently means”
(CAD)$^2$RL: Real Single-Image Flight without a Single Real Image (2016, background): “In [25], laser range scanned real images are used to estimate depth in a supervised learning approach and then the output is used to learn control policies.”
Learning monocular reactive UAV control in cluttered natural environments (2012, background): “Most closely related to our approach are approaches which learn motion policies and depth from input data.”
Target-driven visual navigation in indoor scenes using deep reinforcement learning (2016, methodology): “[27] propose an RL-based method for obstacle detection using a monocular camera.”
Render for CNN: Viewpoint Estimation in Images Using CNNs Trained with Rendered 3D Model Views (2015, methodology): “Synthesizing Images for Training Synthetic images rendered from 3D models have long been used in computer vision [25, 26, 37, 23].”
Safe Visual Navigation via Deep Learning and Novelty Detection (2017, background): “While we focus on supervised learning of collision probability, there are many other ways to map between images and actions, including end-to-end neural networks [6, 19], affordance-based representations [7], reinforcement learning using either simulated or real images [23, 31], and classification o”
Pushbroom stereo for high-speed navigation in cluttered environments (2014, background): “ocular depth estimation, however, is generally not fast and reliable enough for obstacle avoidance on fast-ﬂying UAVs. Machine learning work has shown progress, such as Michels’s radio controlled car [22]. More modern work includes scale-estimation in hover and slow ﬂight regimes [26]. When integrat”
Toward low-flying autonomous MAV trail navigation using deep neural networks for environmental awareness (2017, methodology): “A number of depth-from-single-image techniques have been developed over the years using both traditional machine learning [20], [28] as well as more recent deep learning methods [7], [19].”
Learning Depth from Single Monocular Images (2005, methodology): “Texture information is mostly contained within the image in tensity channel, 2 so we apply Laws’ masks [15, 4] to this channel to compute the texture ene rgy (Fig.”
3-D Depth Reconstruction from a Single Still Image (2007, methodology): “…such as a sunﬂower ﬁeld, an oil-painting scene, mountains and lakes, a city skyline photographed from sea, a city during snowfall, etc. Car Driving Experiments : Michels, Saxena and Ng [20] used a simpliﬁed version of the monocular depth estimation algorithm to drive a remote-controlled car (Fig.”
Learning Transferable Policies for Monocular Reactive MAV Control (2016, background): “However, in recent years learning based methods [15, 16, 13, 11] have become a very powerful alternative to designing hand-engineered perception and control software, even for basic tasks like collision avoidance.”
A Survey of End-to-End Driving: Architectures and Training Methods (2020, background): “in [9]), one might also just explicitly punish crashes by assigning a large negative reward, as was done in [26], [61], [66], [69], [70].”