Accelerate Reinforcement Learning with PID Controllers in the Pendulum Simulations

We propose a Proportional Integral Derivative (PID) controller-based coaching scheme to expedite reinforcement learning (RL).

GNN-PMB: A Simple but Effective Online 3D Multi-Object Tracker without Bells and Whistles

Multi-object tracking (MOT) is among crucial applications in modern advanced driver assistance systems (ADAS) and autonomous driving (AD) systems. Most solutions to MOT are based on random vector Bayesian filters like global nearest neighbor (GNN) plus rule-based heuristical track maintenance. With the development of random finite set (RFS) theory, the RFS Bayesian filters have been applied in MOT tasks for ADAS and AD systems recently. However, their usefulness in the real traffic is open to doubt due to computational cost and implementation complexity. In this paper, it is revealed that GNN with rule-based heuristic track maintenance is insufficient for LiDAR-based MOT tasks in ADAS and AD systems. This judgement is illustrated by systematically comparing several different multi-point object filter-based tracking frameworks, including traditional random vector Bayesian filters with rule-based heuristical track maintenance and RFS Bayesian filters. Moreover, a simple and effective tracker, namely Poisson multi-Bernoulli filter using global nearest neighbor (GNN-PMB) tracker, is proposed for LiDAR-based MOT tasks. The proposed GNN-PMB tracker achieves competitive results in nuScenes test dataset, and shows superior tracking performance over other state-of-the-art LiDAR only trackers and LiDAR and camera fusion-based trackers.

Deep Instance Segmentation with High-Resolution Automotive Radar

Automotive radar has been widely used in the modern advanced driver assistance systems (ADAS) and autonomous driving system as it provides reliable environmental perception in all-weather conditions with affordable cost. However, automotive radar usually only plays as an auxiliary sensor since it hardly supplies semantic and geometry information due to the sparsity of radar detection points. Nonetheless, as development of high-resolution automotive radar in recent years, more advanced perception functionality like instance segmentation which has only been well explored using Lidar point clouds, becomes possible by using automotive radar. Its data comes with rich contexts such as Radar Cross Section (RCS) and micro-doppler effects which may potentially be pertinent, and sometimes can even provide detection when the field of view is completely obscured. Therefore, the effective utilization of radar detection points data is an integral part of automotive perception. The outcome from instance segmentation could be seen as comparable result of clustering, and could be potentially used as the input of tracker for tracking the targets. In this paper, we propose two efficient methods for instance segmentation with radar detection points, one is implemented in an end-to-end deep learning driven fashion using PointNet++ framework, and the other is based on clustering of the radar detection points with semantic information. Both approaches can be further improved by implementing visual multi-layer perceptron (MLP). The effectiveness of the proposed methods is verified using experimental results on the recent RadarScenes dataset.