Learning to infer human attention in daily activities

作者：

Highlights：

• Different encoder-decoder architectures significantly affect the performance.

• The performance improves after considering the task encoding loss.

• The pretrain of the neural network contributes to better performance.

摘要

•Different encoder-decoder architectures significantly affect the performance.•The performance improves after considering the task encoding loss.•The pretrain of the neural network contributes to better performance.

论文关键词：Human attention,Deep neural network,Attentional objects

论文评审过程：Received 10 August 2019, Revised 18 February 2020, Accepted 24 February 2020, Available online 26 February 2020, Version of Record 4 March 2020.

论文官网地址：https://doi.org/10.1016/j.patcog.2020.107314