A Triplet network framework based automatic assessment of simulation quality for respiratory droplet propagation

作者：

Highlights：

• A Triplet network framework with attentive temporal pooling is proposed, which is the first trial to assess the quality of simulation for droplet respiratory propagation automatically.

• A regularization constraint for triplet loss boosts the performance.

• The proposed approach has generalization for other applications such as 3D object retrieval.

摘要

•A Triplet network framework with attentive temporal pooling is proposed, which is the first trial to assess the quality of simulation for droplet respiratory propagation automatically.•A regularization constraint for triplet loss boosts the performance.•The proposed approach has generalization for other applications such as 3D object retrieval.

论文关键词：Simulation quality assessment,Respiratory droplet propagation,Triplet network,Multi-scale CNN-BiLSTM,Attentive temporal pooling

论文评审过程：Received 10 October 2020, Revised 8 May 2021, Accepted 16 May 2021, Available online 8 June 2021, Version of Record 22 June 2021.

论文官网地址：https://doi.org/10.1016/j.patcog.2021.108060