arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1806.08523cs.CV
收藏

专注于相关的东西:使用注意力的时间序列学习和理解

Focusing on What is Relevant: Time-Series Learning and Understanding using Attention

IBM Research(IBM研究院)

Phongtharin Vinayavekhin, Subhajit Chaudhury, Asim Munawar, Don Joven Agravante, Giovanni De Magistris, Daiki Kimura, Ryuki Tachibana

摘要:

本文是对时间序列的不同应用中的深度学习模型的可解释性的贡献。我们提出了一个时间注意层,能够选择相关的信息来执行各种任务,包括数据完成,关键帧检测和分类。该方法使用整个输入序列来计算每个时间步长的注意力值。这导致比以前的方法更集中的注意力值和更合理的可视化。我们将所提出的方法应用于三个不同的任务。实验结果表明,所提出的网络产生可比的结果,以最先进的。此外,该网络提供了更好的可解释性的决定,也就是说,它产生更显着的关注权相关帧相比,类似的技术尝试在过去。

英文摘要:

This paper is a contribution towards interpretability of the deep learning models in different applications of time-series. We propose a temporal attention layer that is capable of selecting the relevant information to perform various tasks, including data completion, key-frame detection and classification. The method uses the whole input sequence to calculate an attention value for each time step. This results in more focused attention values and more plausible visualisation than previous methods. We apply the proposed method to three different tasks. Experimental results show that the proposed network produces comparable results to a state of the art. In addition, the network provides better interpretability of the decision, that is, it generates more significant attention weight to related frames compared to similar techniques attempted in the past.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑