Off-Road Obstacle Avoidance through End-to-End Learning

By Yann LeCun, Urs Muller, Jan Ben, E. Cosatto, B. Flepp. 2005

623 citations recorded by the source corpus.

How researchers see this paper

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

In Robotics & Control

Citations in the corpus

Composable Action-Conditioned Predictors: Flexible Off-Policy Learning for Robot Navigation (2018, background): “Much of the prior work on learning-based robot navigation has focused on imitation learning [19, 20, 21, 22, 23], which requires demonstrations.”
Learning to Drive in a Day (2018, methodology): “Localisation such as [15] facilitates control of the vehicle [16] within the mapped environment, while perception methods such as semantic segmentation [1] enable the robot to interpret the scene.”
End-to-End Driving Via Conditional Imitation Learning (2017, methodology): “This approach has been applied to lane following [27], [4] and off-road obstacle avoidance [22].”
Agile Autonomous Driving using End-to-End Deep Imitation Learning (2017, background): “As a result, high-frequency application of both steering and throttle commands are required in our task, whereas previous work only focuses on steering commands [20, 2, 24].”
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.”
CARLA: An Open Urban Driving Simulator (2017, background): “21,8]. The second approach is based on a deep network trained end-to-end via imitation learning [4]. This approach represents a long line of investigation that has recently attracted renewed interest [22,16,4]. The third approach is based on a deep network trained end-to-end via reinforcement learni”
DeepDriving: Learning Affordance for Direct Perception in Autonomous Driving (2015, background): “In terms of deep learning for autonomous driving, [14] is a successful example of ConvNets-based behavior reflex approach.”
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”
Query-Efficient Imitation Learning for End-to-End Autonomous Driving (2016, methodology): “A similar approach was taken later in 2005 to train, this time, a convolutional neural network to drive an off-road mobile robot [11].”
Imitation Learning via Off-Policy Distribution Matching (2019, background): “This setting describes a number of real-world applications where rewards are unknown, such as Pomerleau (1989); Muller et al. (2006); Bojarski et al. (2016).”
FPGA-Based Accelerators of Deep Learning Networks for Learning and Classification: A Review (2019, background): “image classification [27]–[32], face detection and recognition [33], [34], robot navigation [35]–[37], real-time multiple object tracking [38], financial forecasting [39], and medical diagnosis systems [40]–[42], to name a few.”
Deep semantic segmentation for automated driving: Taxonomy, roadmap and challenges (2017, background): “The work in [13][14] is in that direction.”
Virtual-to-real deep reinforcement learning: Continuous control of mobile robots for mapless navigation (2017, methodology): “For learning-based obstacle avoidance, deep neural networks have been successfully applied on monocular images [11] and depth images [12].”
Learning by Cheating (2019, methodology): “It was pioneered by Pomerleau [19] and developed further in subsequent work [5, 13, 23].”