… a safety module for deep reinforcement learning … safety enhancement for DRL models
without further training or policy updates. This safety module will be useful in real-world safety-…

… whether its future states are safe or one of them leads to a … a deep reinforcement framework
enhanced with a learning-based safety component to achieve a more efficient level of safety …

… a deep reinforcement learning (DRL) approach to enhance the … The proposed strategy aims
to achieve traffic efficiency, safety (… can achieve a good balance between safety, comfort, and …

… It is worth stating that our goal is not only to enhance safety in the solution of the RL algorithm,
but also do so during the training procedure. This is relevant for applications such as self-…

… To address such issues, this paper presents an enhanced BCQ algorithm by … -based safety
enhancement strategy is incorporated to constrain the explorable state space within a safe …

… safety, and driving comfort through dynamic adaptation of the inter-vehicle gap based on deep
reinforcement … and driving safety/comfort by effectively controlling the driving safety model …

… Laine, and MJ Kochenderfer, “Combining planning and deep reinforcement learning in
tactical decision making for autonomous driving,” IEEE Trans. Intell. Vehicles, vol. …

… match and safety when the two differ. Our results for safety-assurance-enhanced variants
of … We believe that online safety assurance can facilitate the safe adoption of DL-augmented …

… ) system for improving the safety of urban expressways in … VSL-based real-time safety
interventions. Existing models are … we employ a deep Q-network, which is a reinforcement …

… learning-based decision-making strategy and analyse the safety … to enhance the
decision-making process. Finally, in the safety efficiency analysis phase, we investigated the safety …