This paper presents a real-time and safe motion planning framework for autonomous driving systems using the Model Predictive Path Integral (MPPI) approach. The proposed method can handle obstacles and guarantee bounds for speed, acceleration, and steering rate to generate feasible and collision-free trajectories.
Current pedestrian detectors exhibit significant bias towards children, highlighting the need for fairness improvements.
Proposing SurroundSDF for accurate and continuous 3D perception using Signed Distance Fields.
Proposing EMIE-MAP for accurate large-scale road surface reconstruction using explicit mesh and implicit encoding.
TrajectoryNAS introduces a pioneering method utilizing point cloud data for trajectory prediction, enhancing the performance of autonomous driving systems.
Proposing EMIE-MAP for large-scale road surface reconstruction using explicit mesh and implicit encoding.
TrajectoryNAS automates the design of trajectory prediction models using Neural Architecture Search, improving accuracy and efficiency in autonomous driving systems.
高精度(HD)マップの効率的な構築方法を提案する。
Introducing SGADS to enhance safety, generalization, and training efficiency in autonomous driving.
The author introduces DyRoNet, a framework that utilizes low-rank dynamic routing to enhance streaming perception in autonomous driving systems. By integrating specialized pre-trained branch networks and a speed router module, DyRoNet achieves a balance between latency and precision.