Deep Reinforcement Learning framework for Autonomous Driving

  title={Deep Reinforcement Learning framework for Autonomous Driving},
  author={Ahmad El Sallab and Mohammed Abdou and Etienne Perot and Senthil Kumar Yogamani},
Reinforcement learning is considered to be a strong AI paradigm which can be used to teach machines through interaction with the environment and learning from their mistakes. [] Key Method It incorporates Recurrent Neural Networks for information integration, enabling the car to handle partially observable scenarios. It also integrates the recent work on attention models to focus on relevant information, thereby reducing the computational complexity for deployment on embedded hardware. The framework was…

Figures and Tables from this paper

Deep Reinforcement Learning Framework for Navigation in Autonomous Driving

The purpose of this work is to implement navigation in autonomous car using MXNet, an open source reinforcement learning framework which is primarily used to train and deploy deep neural networks.

Learning to Drive with Deep Reinforcement Learning

End-to-end reinforcement learning for an autonomous car is described, which used only a single reinforcement learning model to create the autonomous car and it is shown that it performed better with state-of-the-art off-policy reinforcementlearning for continuous action (SAC, TD3).

Review of Deep Reinforcement Learning for Autonomous Driving

This research presents a nomenclature of autonomous driving in which DRL techniques have been used, thus discussing important computational issues in evaluating autonomous driving agents in the real environment.

Multi-Agent Connected Autonomous Driving using Deep Reinforcement Learning

  • Praveen Palanisamy
  • Computer Science
    2020 International Joint Conference on Neural Networks (IJCNN)
  • 2020
This paper proposes the use of Partially Observable Markov Games (POSG) for formulating the connected autonomous driving problems with realistic assumptions and provides a taxonomy of multi-agent learning environments based on the nature of tasks, nature of agents and thenature of the environment to help in categorizing various autonomousdriving problems that can be addressed under the proposed formulation.

Combining YOLO and Deep Reinforcement Learning for Autonomous Driving in Public Roadworks Scenarios

Deep Deterministic Policy Gradient (DDPG) is used for controlling the behavior of the vehicle when overtaking the working zones and YOLOv3-tiny achieved an 94.8% mAP and proved to be reliable in real-world applications.

Deep Hierarchical Reinforcement Learning for Autonomous Driving with Distinct Behaviors

This paper designs a hierarchical neural network policy and proposes a hierarchical policy gradient method to train the network with the semi markov decision process (SMDP) temporal abstraction formulation and shows via simulation that this method is able to select correct decision and acts appropriately when the traffic light turns yellow.

Hierarchical Program-Triggered Reinforcement Learning Agents For Automated Driving

HPRL - Hierarchical Program-triggered Reinforcement Learning is proposed, which uses a hierarchy consisting of a structured program along with multiple RL agents, each trained to perform a relatively simple task, leading to a significantly more interpretable and verifiable implementation as compared to a complex RL agent.

Deep Distributional Reinforcement Learning Based High-Level Driving Policy Determination

A supervisor agent that can enhance the driver assistant systems by using deep distributional reinforcement learning is proposed, trained using end-to-end approach that directly maps both a camera image and LIDAR data into action plan.


This paper presents the state of the art in deep reinforcement learning paradigm highlighting the current achievements for autonomous driving vehicles.

Online Virtual Training in Soft Actor-Critic for Autonomous Driving

  • Maryam SavariY. Choe
  • Computer Science
    2021 International Joint Conference on Neural Networks (IJCNN)
  • 2021
This research pre-trained Soft Actor-Critic with Learning from Demonstrations to find out if pre-training can reduce the training time of the SAC algorithm and proposed online end-to-end combination method of SAC, LfD, and Learning from Interventions is proposed to train an agent.



Human-level control through deep reinforcement learning

This work bridges the divide between high-dimensional sensory inputs and actions, resulting in the first artificial agent that is capable of learning to excel at a diverse array of challenging tasks.

An Empirical Evaluation of Deep Learning on Highway Driving

It is shown how existing convolutional neural networks can be used to perform lane and vehicle detection while running at frame rates required for a real-time system, lending credence to the hypothesis that deep learning holds promise for autonomous driving.

Deep Recurrent Q-Learning for Partially Observable MDPs

The effects of adding recurrency to a Deep Q-Network is investigated by replacing the first post-convolutional fully-connected layer with a recurrent LSTM, which successfully integrates information through time and replicates DQN's performance on standard Atari games and partially observed equivalents featuring flickering game screens.

Playing Atari with Deep Reinforcement Learning

This work presents the first deep learning model to successfully learn control policies directly from high-dimensional sensory input using reinforcement learning, which outperforms all previous approaches on six of the games and surpasses a human expert on three of them.

Deep Attention Recurrent Q-Network

Tests of the proposed Deep Attention Recurrent Q-Network (DARQN) algorithm on multiple Atari 2600 games show level of performance superior to that of DQN.

End to End Learning for Self-Driving Cars

A convolutional neural network is trained to map raw pixels from a single front-facing camera directly to steering commands and it is argued that this will eventually lead to better performance and smaller systems.

Reinforcement Learning: An Introduction

This book provides a clear and simple account of the key ideas and algorithms of reinforcement learning, which ranges from the history of the field's intellectual foundations to the most recent developments and applications.

Deep Tracking: Seeing Beyond Seeing Using Recurrent Neural Networks

This paper presents to the best of our knowledge the first end-to-end object tracking approach which directly maps from raw sensor input to object tracks in sensor space without requiring any

Simple statistical gradient-following algorithms for connectionist reinforcement learning

This article presents a general class of associative reinforcement learning algorithms for connectionist networks containing stochastic units that are shown to make weight adjustments in a direction that lies along the gradient of expected reinforcement in both immediate-reinforcement tasks and certain limited forms of delayed-reInforcement tasks, and they do this without explicitly computing gradient estimates.

Recurrent Models of Visual Attention

A novel recurrent neural network model that is capable of extracting information from an image or video by adaptively selecting a sequence of regions or locations and only processing the selected regions at high resolution is presented.