Scholarlink

Event segmentation in temporal data [1] is crucial for cognitive architectures and diverse applications. This work proposes an innovative model, inspired by neuroscience and psychology, that emulates the human brain’s temporal segmentation during visual perception. It integrates perceptual processes with hierarchical attention to detect event transitions in dynamic environments, generating experience-based knowledge for memory, planning, and decision-making. Evaluated in a real-life visual scenario, the model accurately identified events even in noisy conditions, offering a promising framework for adaptive cognitive systems, including autonomous robotics.

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

Bio-Inspired Computational Model of Perception for the Detection of Simple Events from Their Visual Features

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Alison Muñoz Capote,
  • Victor Manuel Pelaez Fernandez,
  • Francisco Ismael Lopez Gomez,
  • Marco Antonio Ramos Corchado,
  • Jose Antonio Cervantes,
  • Félix Francisco Ramos Corchado

摘要

Event segmentation in temporal data [1] is crucial for cognitive architectures and diverse applications. This work proposes an innovative model, inspired by neuroscience and psychology, that emulates the human brain’s temporal segmentation during visual perception. It integrates perceptual processes with hierarchical attention to detect event transitions in dynamic environments, generating experience-based knowledge for memory, planning, and decision-making. Evaluated in a real-life visual scenario, the model accurately identified events even in noisy conditions, offering a promising framework for adaptive cognitive systems, including autonomous robotics.