Title: A reinforcement learning intelligent deductive model with pre-trained sequence information

Authors: Xinyu Han; Huosheng Xu; Hao Yu; Sizhao Li

Addresses: College of Computer Science and Technology, Harbin Engineering University, Harbin, Heilongjiang Province, China ' College of Computer Science and Technology, Harbin Engineering University, Harbin, Heilongjiang Province, China; Wuhan Digital Engineering Institute, Wuhan, Hubei Province, China ' College of Computer Science and Technology, Harbin Engineering University, Harbin, Heilongjiang Province, China ' College of Computer Science and Technology, Harbin Engineering University, Harbin, Heilongjiang Province, China

Abstract: Agent trajectory prediction is an increasingly popular topic in computer vision and autonomous driving. With the help of deep learning and big data, it is possible to understand the interaction model between agents hidden in complex environments. Existing methods usually pay more attention to the average trajectory offset of the agent while ignoring the distribution differences of the target. This issue results inevitable performance decrease. To address this issue, we propose a novel reinforcement learning intelligent deduction model (RLDM). It achieves joint reasoning of goals and paths in a unified framework, and accurately predicts trajectories in a short period of time with fewer datasets. Specifically, an end-to-end time-series pre-training module is proposed to explore the agent's training state reward and goal reward. Moreover, a prediction module based on the combination of kinematics and environmental background is proposed to explore the agent motion characteristics. By this way, acting in a purely reactive manner is better relieved. Practical trajectory prediction experiments are designed, and the experimental results verify the superior performance of our proposed model. The model experiment results are improved by 2% and 11% on the ADE and FDE metric on average.

Keywords: reinforcement learning; trajectory prediction; intelligent deduction; neural networks.

DOI: 10.1504/IJBIC.2023.136098

International Journal of Bio-Inspired Computation, 2023 Vol.22 No.4, pp.195 - 205

Received: 30 Jun 2022
Accepted: 09 Jun 2023
Published online: 16 Jan 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article