Vision-Based Obstacle Avoidance for UAVs via Imitation Learning with Sequential Neural Networks

Bumsoo, Park; Oh, Hyondong

doi:10.1007/s42405-020-00254-x

Cited by 18 publications

(8 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Imitation learning has also seen several applications in the military domain [58,59]. For example in [60], imitation learning is applied to learning decision policies of computer generated forces.…”

Section: Imitating the Military Expertmentioning

confidence: 99%

Data-driven behavioural modelling for military applications

Schadd¹,

Reus²,

Uilkema³

et al. 2022

JDST

View full text Add to dashboard Cite

This article investigates the possibilities for creating behavioural models of military decision making in a data-driven manner. As not much data from actual operations is available, and data cannot easily be created in the military context, most approaches use simulators to learn behaviour. A simulator is however not always available or is difficult to create. This study focusses on the creation of behavioural models from data that was collected during a field exercise. As data in general is limited, noisy and erroneous, this makes the creation of realistic models challenging. Besides using the traditional approach of hand-crafting a model based on data, we investigate the emerging research area of imitation learning. One of its techniques, reward engineering, is applied to learn the behaviour of soldiers in an urban warfare operation. Basic, but realistic, soldier behaviour is learned, which lays the groundwork for more elaborate models in the future.

show abstract

Section: Imitating the Military Expertmentioning

confidence: 99%

Data-driven behavioural modelling for military applications

Schadd¹,

Reus²,

Uilkema³

et al. 2022

JDST

View full text Add to dashboard Cite

show abstract

“…In order to solve the obstacle avoidance problem of UAV visual navigation, Park et al [124] proposed an imitation learning strategy based on human experts sharing flight data. As shown in FIGURE 25, the strategy takes UAV visual image as input, uses neural network to extract features, and trains UAV obstacle avoidance strategy.…”

Section: (A) (B) Figure24 Navigation In a Complex Environmentmentioning

confidence: 99%

Motion Planning for Mobile Robots—Focusing on Deep Reinforcement Learning: A Systematic Review

Sun

Zhang

et al. 2021

IEEE Access

View full text Add to dashboard Cite

Mobile robots contributed significantly to the intelligent development of human society, and the motion-planning policy is critical for mobile robots. This paper reviews the methods based on motionplanning policy, especially the ones involving Deep Reinforcement Learning (DRL) in the unstructured environment. The conventional methods of DRL are categorized to value-based, policy-based and actorcritic-based algorithms, and the corresponding theories and applications are surveyed. Furthermore, the recently-emerged methods of DRL are also surveyed, especially the ones involving the imitation learning, meta-learning and multi-robot systems. According to the surveys, the potential research directions of motion-planning algorithms serving for mobile robots are enlightened.

show abstract

“…The recurrent network established the effective avoidance of static and dynamic obstacles. Recently, a vision-based obstacle avoidance approach was developed based on a sequential NN [46]. The study was focused on an aerospace application but similar models can be adopted for flexible robotic surgery.…”

Section: Literature Reviewmentioning

confidence: 99%

Motion and Trajectory Constraints Control Modeling for Flexible Surgical Robotic Systems

et al. 2020

View full text Add to dashboard Cite

Success of the da Vinci surgical robot in the last decade has motivated the development of flexible access robots to assist clinical experts during single-port interventions of core intrabody organs. Prototypes of flexible robots have been proposed to enhance surgical tasks, such as suturing, tumor resection, and radiosurgery in human abdominal areas; nonetheless, precise constraint control models are still needed for flexible pathway navigation. In this paper, the design of a flexible snake-like robot is presented, along with the constraints model that was proposed for kinematics and dynamics control, motion trajectory planning, and obstacle avoidance during motion. Simulation of the robot and implementation of the proposed control models were done in Matlab. Several points on different circular paths were used for evaluation, and the results obtained show the model had a mean kinematic error of 0.37 ± 0.36 mm with very fast kinematics and dynamics resolution times. Furthermore, the robot's movement was geometrically and parametrically continuous for three different trajectory cases on a circular pathway. In addition, procedures for dynamic constraint and obstacle collision detection were also proposed and validated. In the latter, a collision-avoidance scheme was kept optimal by keeping a safe distance between the robot's links and obstacles in the workspace. Analyses of the results showed the control system was optimal in determining the necessary joint angles to reach a given target point, and motion profiles with a smooth trajectory was guaranteed, while collision with obstacles were detected a priori and avoided in close to real-time. Furthermore, the complexity and computational effort of the algorithmic models were negligibly small. Thus, the model can be used to enhance the real-time control of flexible robotic systems.

show abstract

Vision-Based Obstacle Avoidance for UAVs via Imitation Learning with Sequential Neural Networks

Cited by 18 publications

References 17 publications

Data-driven behavioural modelling for military applications

Data-driven behavioural modelling for military applications

Motion Planning for Mobile Robots—Focusing on Deep Reinforcement Learning: A Systematic Review

Motion and Trajectory Constraints Control Modeling for Flexible Surgical Robotic Systems

Contact Info

Product

Resources

About