Perception is crucial for drone obstacle avoidance in complex, static, and unstructured outdoor environments. However, most navigation solutions based on Deep Reinforcement Learning (DRL) use limited Field-Of-View (FOV) images as input. In this paper, we demonstrate that omnidirectional improve these methods. Thus, provide a comparative benchmark of several visual modalities navigation: ground ...

This paper aims at highlighting cutting-edge research results in the field of visual tracking by deep reinforcement learning. Deep learning (DRL) is an emerging area combining recent progress and It showing interesting computer vision and, recently, it has been applied to problem yielding rapid development novel strategies. After providing introduction learning, this compares approaches based o...

Deep reinforcement learning (DRL), combining the perception capability of deep (DL) and decision-making (RL) [1], has been widely investigated for autonomous driving tasks. In this letter, we would like to discuss impact different types state input on performance DRL-based lane change decision-making.

Deep reinforcement learning (DRL) algorithms have recently gained wide attention in the wireless networks domain. They are considered promising approaches for solving dynamic radio resource management (RRM) problems next-generation networks. Given their capabilities to build an approximate and continuously updated model of network environments, DRL can deal with multifaceted complexity such env...

With the reform of energy structures, high proportion volatile new access makes existing unit commitment (UC) theory unable to satisfy development demands day-ahead market decision-making in power system. Therefore, this paper proposes an intelligent algorithm for solving UC, based on deep reinforcement learning (DRL) technology. Firstly, DRL is used model Markov decision process UC problem, an...

Deep Reinforcement Learning (DRL) has shown a dramatic improvement in decision-making and automated control problems. Consequently, DRL represents promising technique to efficiently solve many relevant optimization problems (e.g., routing) self-driving networks. However, existing DRL-based solutions applied networking fail generalize, which means that they are not able operate properly when net...