This chapter intends to provide some directions for future research in the crucial endeavor of understanding and modeling human attention from both physiological and computer science perspectives. In addition to only providing new attention models, the chapter provides some cues on the role of attention towards autonomy and a more general and “curious” AI. In that perspective, we stress the importance of working on the convergence of the different communities needed to achieve links between deep learning, artificial attention, and human or animal attention implying computer science, machine learning, computational attention, cognitive psychology, and neuroscience. A better integration of those different domains is crucial for both a better understanding of machines but also of ourselves.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

The Future of Attention Models: Convergence of Deep Learning with Artificial and Human Attention

  • Matei Mancas,
  • Vincent P. Ferrera,
  • Antoine Coutrot

摘要

This chapter intends to provide some directions for future research in the crucial endeavor of understanding and modeling human attention from both physiological and computer science perspectives. In addition to only providing new attention models, the chapter provides some cues on the role of attention towards autonomy and a more general and “curious” AI. In that perspective, we stress the importance of working on the convergence of the different communities needed to achieve links between deep learning, artificial attention, and human or animal attention implying computer science, machine learning, computational attention, cognitive psychology, and neuroscience. A better integration of those different domains is crucial for both a better understanding of machines but also of ourselves.