Transportation Service Redundancy From a Spatio-Temporal Perspective

Haddad, Hedi; Bouyahia, Zied; Jabeur, Nafaâ

doi:10.1109/mits.2019.2939139

Cited by 2 publications

(4 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the future, the driving scenario is expected to emulate more real-world driving complexities, extended to multiple vehicles in the driving lane, vehicles on the oncoming lane, traffic lights, intersections, and curvatures in the driving trajectory. Our results outperform some data-driven models based on the accuracy, maximum error-free drive-time before deviating off the track and policy losses [ 14 ]. The state-of-the-art developments in DL and RL were the principal motivation behind this work.…”

Section: Introductionmentioning

confidence: 80%

Policy-Gradient and Actor-Critic Based State Representation Learning for Safe Driving of Autonomous Vehicles

Gupta

Khwaja

Anpalagan

et al. 2020

Sensors

View full text Add to dashboard Cite

In this paper, we propose an environment perception framework for autonomous driving using state representation learning (SRL). Unlike existing Q-learning based methods for efficient environment perception and object detection, our proposed method takes the learning loss into account under deterministic as well as stochastic policy gradient. Through a combination of variational autoencoder (VAE), deep deterministic policy gradient (DDPG), and soft actor-critic (SAC), we focus on uninterrupted and reasonably safe autonomous driving without steering off the track for a considerable driving distance. Our proposed technique exhibits learning in autonomous vehicles under complex interactions with the environment, without being explicitly trained on driving datasets. To ensure the effectiveness of the scheme over a sustained period of time, we employ a reward-penalty based system where a negative reward is associated with an unfavourable action and a positive reward is awarded for favourable actions. The results obtained through simulations on DonKey simulator show the effectiveness of our proposed method by examining the variations in policy loss, value loss, reward function, and cumulative reward for `VAE+DDPG’ and `VAE+SAC’ over the learning process.

show abstract

Section: Introductionmentioning

confidence: 80%

Policy-Gradient and Actor-Critic Based State Representation Learning for Safe Driving of Autonomous Vehicles

Gupta

Khwaja

Anpalagan

et al. 2020

Sensors

View full text Add to dashboard Cite

show abstract

“…POMDPs formulate the autonomous vehicle control problem as an optimization task, and rely on assumptions to optimize an objective [42]. The RL seems to be promising for planning and control aspects and scales to very complex environments and unexpected scenarios [44].…”

Section: Empirical Decision-making System For Autonomous Vehiclesmentioning

confidence: 99%

“…When a regularization factor is applied, the DL model learns a transformation while still adjusting the reward function and policy loss due to correlations in time [58]. Previous approaches available in literature have learned the function F directly, in case of processing artificial videos [44]. The function F is learned in a piecewise manner so that the efficiency and performance can be improved separately [78].…”

Section: Scenario Setup In Donkey Simulatormentioning

confidence: 99%

“…The autonomous driving problem is partly restricted to a learning problem constrained in the pixel space defined as a hypersphere of radius ρ [44], [54], [79]. To ameliorate the problem complexity, we decided to learn action prediction with a policy gradient architecture (DDPG) and an actor-critic architecture (SAC), both preceded by an autoencoder:…”

Section: Scenario Setup In Donkey Simulatormentioning

confidence: 99%

See 1 more Smart Citation

Safe Driving Of Autonomous Vehicles Through Improved Deep Reinforcement Learning

Gupta¹

2021

Preprint

View full text Add to dashboard Cite

In this thesis, we propose an environment perception framework for autonomous driving using deep reinforcement learning (DRL) that exhibits learning in autonomous vehicles under complex interactions with the environment, without being explicitly trained on driving datasets. Unlike existing techniques, our proposed technique takes the learning loss into account under deterministic as well as stochastic policy gradient. We apply DRL to object detection and safe navigation while enhancing a self-driving vehicle’s ability to discern meaningful information from surrounding data. For efficient environmental perception and object detection, various Q-learning based methods have been proposed in the literature. Unlike other works, this thesis proposes a collaborative deterministic as well as stochastic policy gradient based on DRL. Our technique is a combination of variational autoencoder (VAE), deep deterministic policy gradient (DDPG), and soft actor-critic (SAC) that adequately trains a self-driving vehicle. In this work, we focus on uninterrupted and reasonably safe autonomous driving without colliding with an obstacle or steering off the track. We propose a collaborative framework that utilizes best features of VAE, DDPG, and SAC and models autonomous driving as partly stochastic and partly deterministic policy gradient problem in continuous action space, and continuous state space. To ensure that the vehicle traverses the road over a considerable period of time, we employ a reward-penalty based system where a higher negative penalty is associated with an unfavourable action and a comparatively lower positive reward is awarded for favourable actions. We also examine the variations in policy loss, value loss, reward function, and cumulative reward for ‘VAE+DDPG’ and ‘VAE+SAC’ over the learning process.

show abstract

Transportation Service Redundancy From a Spatio-Temporal Perspective

Cited by 2 publications

References 18 publications

Policy-Gradient and Actor-Critic Based State Representation Learning for Safe Driving of Autonomous Vehicles

Policy-Gradient and Actor-Critic Based State Representation Learning for Safe Driving of Autonomous Vehicles

Safe Driving Of Autonomous Vehicles Through Improved Deep Reinforcement Learning

Contact Info

Product

Resources

About