Sparse motion fields for trajectory prediction

作者：

Highlights：

• A model based on sparse motion fields is proposed to characterize motion patterns in trajectory data.

• The model is able separately identify the multiple kinds of motion in a video scene.

• The model automatically learns cues about the static environment (e.g., walkable areas or obstacles) solely from trajectory data.

• The experimental results in several benchmarking data sets show that the proposed approach is competitive with the state of the art in the long-term trajectory prediction task.

摘要

•A model based on sparse motion fields is proposed to characterize motion patterns in trajectory data.•The model is able separately identify the multiple kinds of motion in a video scene.•The model automatically learns cues about the static environment (e.g., walkable areas or obstacles) solely from trajectory data.•The experimental results in several benchmarking data sets show that the proposed approach is competitive with the state of the art in the long-term trajectory prediction task.

论文关键词：Human motion analysis,Trajectory prediction,Sparse motion fields

论文评审过程：Received 22 January 2020, Revised 9 August 2020, Accepted 6 September 2020, Available online 10 September 2020, Version of Record 14 September 2020.

论文官网地址：https://doi.org/10.1016/j.patcog.2020.107631