AC-VRNN: Attentive Conditional-VRNN for multi-future trajectory prediction

作者：

Highlights：

• Multi-future trajectory predictions in crowded scenarios are considered.

• We propose a model based on Conditional Variational Recurrent Neural Networks.

• Prior belief maps steer predictions mimicking human behaviours.

• An attentive-based graph neural network models interactions among pedestrians.

• We outperform state-of-the-art methods on several trajectory prediction benchmarks.

摘要

•Multi-future trajectory predictions in crowded scenarios are considered.•We propose a model based on Conditional Variational Recurrent Neural Networks.•Prior belief maps steer predictions mimicking human behaviours.•An attentive-based graph neural network models interactions among pedestrians.•We outperform state-of-the-art methods on several trajectory prediction benchmarks.

论文关键词：Trajectory forecasting,Multi-future prediction,Time series,Variational recurrent neural networks,Graph attention networks

论文评审过程：Received 24 July 2020, Revised 24 June 2021, Accepted 30 June 2021, Available online 7 July 2021, Version of Record 15 July 2021.

论文官网地址：https://doi.org/10.1016/j.cviu.2021.103245