Human-Aligned Longitudinal Control for Occluded Pedestrian Crossing With Visual Attention
Authors:
- 13 May 2024
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 7419-7425
Abstract
Reinforcement Learning (RL) has been widely used to create generalizable autonomous vehicles. However, they rely on fixed reward functions that struggle to balance values like safety and efficiency. How can autonomous vehicles balance different driving objectives and human values in a constantly changing environment? To bridge this gap, we propose an adaptive reward function that utilizes visual attention maps to detect pedestrians in the driving scene and dynamically switch between prioritizing safety or efficiency depending on the current observation. The visual attention map is used to provide spatial attention to the RL agent to boost the training efficiency of the pipeline. We evaluate the pipeline against variants of an occluded pedestrian crossing scenario in the CARLA Urban Driving simulator. Specifically, the proposed pipeline is compared against a modular setup that combines the well-established object detection model, YOLO, with a Proximal Policy Optimization (PPO) agent. The results indicate that the proposed approach can compete with the modular setup while yielding greater training efficiency. The trajectories collected with the approach confirm the effectiveness of the proposed adaptive reward function.Keywords
This publication has 22 references indexed in Scilit:
- Explainable Artificial Intelligence for Autonomous Driving: A Comprehensive Overview and Field Guide for Future Research DirectionsIEEE Access, 2024
- Deep Reinforcement Learning for Autonomous Vehicle Intersection Navigation2023
- An efficient planning method based on deep reinforcement learning with hybrid actions for autonomous driving on highwayInternational Journal of Machine Learning and Cybernetics, 2023
- Reinforcement Learning Driving Strategy based on Auxiliary Task for Multi-Scenarios Autonomous Driving2023
- Attention mechanisms in computer vision: A surveyComputational Visual Media, 2022
- Deep Reinforcement Learning-Based Driving Strategy for Avoidance of Chain Collisions and Its Safety Efficiency Analysis in Autonomous VehiclesIEEE Access, 2022
- Scaled-YOLOv4: Scaling Cross Stage Partial Network2021
- Designing AI for Explainability and Verifiability: A Value Sensitive Design Approach to Avoid Artificial Stupidity in Autonomous VehiclesInternational Journal of Social Robotics, 2021
- Grounding Human-To-Vehicle Advice for Self-Driving Vehicles2019
- Deep Residual Learning for Image Recognition2016