Random curiosity-driven exploration in deep reinforcement learning

Li, Jing; Shi, Xinxin; Li, Jiehao; Zhang, Xin; Wang, Junzheng

doi:10.1016/j.neucom.2020.08.024

Cited by 56 publications

(17 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Directed exploration strategies utilize the previous history of the learning process and influence the portion of the environment explored in the future, including count-based [32], curiosity-driven [31,33], and upper confidence bounds (UCB) exploration [34]. For tabular-based RL, count-based exploration strategies give an extra exploration bonus to frequently visited states [32].…”

Section: B Related Workmentioning

confidence: 99%

Rule-Based Reinforcement Learning for Efficient Robot Navigation With Space Reduction

Zhu

Wang

Chen

et al. 2022

IEEE/ASME Trans. Mechatron.

View full text Add to dashboard Cite

For real-world deployments, it is critical to allow robots to navigate in complex environments autonomously. Traditional methods usually maintain an internal map of the environment, and then design several simple rules, in conjunction with a localization and planning approach, to navigate through the internal map. These approaches often involve a variety of assumptions and prior knowledge. In contrast, recent reinforcement learning (RL) methods can provide a model-free, self-learning mechanism as the robot interacts with an initially unknown environment, but are expensive to deploy in realworld scenarios due to inefficient exploration. In this paper, we focus on efficient navigation with the RL technique and combine the advantages of these two kinds of methods into a rulebased RL (RuRL) algorithm for reducing the sample complexity and cost of time. First, we use the rule of wall-following to generate a closed-loop trajectory. Second, we employ a reduction rule to shrink the trajectory, which in turn effectively reduces the redundant exploration space. Besides, we give the detailed theoretical guarantee that the optimal navigation path is still in the reduced space. Third, in the reduced space, we utilize the Pledge rule to guide the exploration strategy for accelerating the RL process at the early stage. Experiments conducted on real robot navigation problems in hex-grid environments demonstrate that RuRL can achieve improved navigation performance.

show abstract

Section: B Related Workmentioning

confidence: 99%

Rule-Based Reinforcement Learning for Efficient Robot Navigation With Space Reduction

Zhu

Wang

Chen

et al. 2022

IEEE/ASME Trans. Mechatron.

View full text Add to dashboard Cite

show abstract

“…A growing number of researchers are applying DRL to solve the perception-decision problem in computer vision tasks [23]. [24] applies a policy gradient based on random curiosity-driven exploration to image coding, resulting in good performance in Atari games. Uzkent et al select HR image blocks with policy gradients to improve classification accuracy with LR images as states [25].…”

Section: B Computer Vision With Deep Reinforcement Learningmentioning

confidence: 99%

Task-Oriented Image Transmission for Scene Classification in Unmanned Aerial Systems

Xu¹,

Song²,

Guo³

et al. 2021

Preprint

View full text Add to dashboard Cite

The vigorous developments of Internet of Things make it possible to extend its computing and storage capabilities to computing tasks in the aerial system with collaboration of cloud and edge, especially for artificial intelligence (AI) tasks based on deep learning (DL). Collecting a large amount of image/video data, Unmanned aerial vehicles (UAVs) can only handover intelligent analysis tasks to the back-end mobile edge computing (MEC) server due to their limited storage and computing capabilities. How to efficiently transmit the most correlated information for the AI model is a challenging topic. Inspired by the task-oriented communication in recent years, we propose a new aerial image transmission paradigm for the scene classification task. A lightweight model is developed on the frontend UAV for semantic blocks transmission with perception of images and channel conditions. In order to achieve the tradeoff between transmission latency and classification accuracy, deep reinforcement learning (DRL) is used to explore the semantic blocks which have the best contribution to the back-end classifier under various channel conditions. Experimental results show that the proposed method can significantly improve classification accuracy compared to the fixed transmission strategy and traditional content perception methods.

show abstract

“…On the other hand, the disadvantage of the above path tracking methods is that all methods believe the robot model and external environment are ideal 21‐23 . Aiming at uncertain disturbances of complex robot systems, common methods include sliding mode control (SMC), active disturbance rejection control (ADRC), symplectic pseudospectral RHC controller, and adaptive control 24‐26 .…”

Section: Introductionmentioning

confidence: 99%

Neural networks‐based sliding mode tracking control for the four wheel‐legged robot under uncertain interaction

Wang

2021

Intl J Robust & Nonlinear

Self Cite

View full text Add to dashboard Cite

When considering the accuracy of tracking control, physical interaction such as structural uncertainties and external dynamics is the main challenge in actual engineering scenarios, especially for the complex robot system. In this article, a neural network‐based sliding mode tracking control scheme (SMCR) is presented for the developed four wheel‐legged robot (BIT‐NAZA) under the uncertain interaction. First, a non‐singular fast terminal function based on the kinematic model is proposed for path tracking, which improves the motion quality during the approach movement and the sliding mode movement. At the same time, it can reduce the influence of uncertain disturbances on the premise of ensuring the path tracking control accuracy via neural networks. Finally, some demonstrations using the autonomous platform of the BIT‐NAZA robot are employed to evaluate the robustness and effectiveness of the hybrid algorithm.

show abstract

Random curiosity-driven exploration in deep reinforcement learning

Cited by 56 publications

References 25 publications

Rule-Based Reinforcement Learning for Efficient Robot Navigation With Space Reduction

Rule-Based Reinforcement Learning for Efficient Robot Navigation With Space Reduction

Task-Oriented Image Transmission for Scene Classification in Unmanned Aerial Systems

Neural networks‐based sliding mode tracking control for the four wheel‐legged robot under uncertain interaction

Contact Info

Product

Resources

About