Advancements in Autonomous Systems and Reinforcement Learning

The field of autonomous systems and reinforcement learning is rapidly evolving, with a focus on developing more efficient and safe decision-making frameworks. Recent developments have centered around improving exploration in complex environments, enhancing the ability of agents to interact with their surroundings, and advancing the state-of-the-art in areas such as autonomous driving and multi-agent systems. Notably, researchers have been exploring the application of reinforcement learning to real-world problems, including traffic management and autonomous vehicle control.

A key area of innovation is the integration of reinforcement learning with other techniques, such as inverse reinforcement learning and graph-based methods, to improve the efficiency and effectiveness of autonomous systems. Additionally, there is a growing emphasis on developing more scalable and reliable multi-agent reinforcement learning frameworks, which can handle complex scenarios and large-scale networks.

Some noteworthy papers in this area include: Dual-Objective Reinforcement Learning with Novel Hamilton-Jacobi-Bellman Formulations, which proposes a new approach to dual-objective reinforcement learning. M-Predictive Spliner: Enabling Spatiotemporal Multi-Opponent Overtaking for Autonomous Racing, which presents a method for autonomous racing that can handle multiple opponents. ParkFormer: A Transformer-Based Parking Policy with Goal Embedding and Pedestrian-Aware Control, which introduces a novel framework for autonomous parking using transformer-based architectures.

Sources

Dual-Objective Reinforcement Learning with Novel Hamilton-Jacobi-Bellman Formulations

M-Predictive Spliner: Enabling Spatiotemporal Multi-Opponent Overtaking for Autonomous Racing

Towards Emergency Scenarios: An Integrated Decision-making Framework of Multi-lane Platoon Reorganization

Improved Exploration in GFlownets via Enhanced Epistemic Neural Networks

Goal-conditioned Hierarchical Reinforcement Learning for Sample-efficient and Safe Autonomous Driving at Intersections

BIDA: A Bi-level Interaction Decision-making Algorithm for Autonomous Vehicles in Dynamic Traffic Scenarios

ParkFormer: A Transformer-Based Parking Policy with Goal Embedding and Pedestrian-Aware Control

Scalable and Reliable Multi-agent Reinforcement Learning for Traffic Assignment

Reinforcement Learning-Based Dynamic Grouping for Tubular Structure Tracking

Path Learning with Trajectory Advantage Regression

Temporal-IRL: Modeling Port Congestion and Berth Scheduling with Inverse Reinforcement Learning

GoIRL: Graph-Oriented Inverse Reinforcement Learning for Multimodal Trajectory Prediction

Evaluation of Traffic Signals for Daily Traffic Pattern

Built with on top of