Hierarchical generator of tracking global hypotheses

作者：

Highlights：

• A Generator of Global Hypotheses that implicitly neglects improbable assignments.

• Association solution in an affordable time.

• Formulation of hypotheses that regard the apparition of new individuals.

• Specialized treatment of the tracks according to their particular characteristics.

• Pre-trained multi-shot neural model to measure appearance affinity.

摘要

•A Generator of Global Hypotheses that implicitly neglects improbable assignments.•Association solution in an affordable time.•Formulation of hypotheses that regard the apparition of new individuals.•Specialized treatment of the tracks according to their particular characteristics.•Pre-trained multi-shot neural model to measure appearance affinity.

论文关键词：Multi-object tracking,Hierarchical data association,Tracking global hypothesis,Appearance neural model

论文评审过程：Received 15 October 2020, Revised 2 June 2022, Accepted 6 June 2022, Available online 8 June 2022, Version of Record 16 June 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.117813