计算机应用

• 人工智能与仿真 •    下一篇

注意力机制综述

任欢,王旭光   

  1. 华北电力大学自动化系
  • 收稿日期:2020-10-21 修回日期:2021-01-07 发布日期:2021-01-07 出版日期:2021-01-27
  • 通讯作者: 任欢

Review of attention mechanism

  • Received:2020-10-21 Revised:2021-01-07 Online:2021-01-07 Published:2021-01-27

摘要: 现在注意力机制已广泛地应用在深度学习的诸多领域。基于注意力机制的结构模型不仅能够记录信息间的位置关系,还能依据信息的权重去度量不同信息特征的重要性。通过对信息特征进行相关与不相关的抉择建立动态权重参数,以加强关键信息弱化无用信息,从而提高深度学习算法效率同时也改进了传统深度学习的一些缺陷。因此从图像处理领域、自然语言处理、数据预测等不同应用方面介绍了一些与注意力机制结合的算法结构,并对近几年大火的基于注意力机制的transformer和reformer算法进行了综述。鉴于注意力机制的重要性,综述了注意力机制的研究发展,分析了注意力机制目前的发展现状并探讨了该机制未来可行的研究方向。

Abstract: Now the attention mechanism has been widely used in many fields of deep learning. The structural model based on the attention mechanism can not only record the positional relationship between information, but also measure the importance of different information features based on the weight of the information. Through the selection of relevant and irrelevant information features, dynamic weight parameters are established to strengthen key information and weaken useless information, thereby improving the efficiency of deep learning algorithms and improving some of the defects of traditional deep learning. Therefore, some algorithm structures combined with the attention mechanism are introduced from different applications in the field of image processing, natural language processing, and data prediction, and the attention mechanism-based transformer and reformer algorithms that have been popular in recent years are reviewed. In view of the importance of attention mechanism, the research development of attention mechanism is reviewed, the current development status of attention mechanism is analyzed, and the feasible research directions of this mechanism in the future are discussed.

中图分类号: