中文

基于视频分类的混合关键事件上下文感知检测

计算机视觉与模式识别 2025-01-03 v2

摘要

通过计算机视觉检测混合关键事件具有挑战性,因为需要理解上下文才能准确评估事件的严重程度。混合关键事件,如不同严重程度的火灾或交通事故,要求具有适应性系统,能够解释上下文以触发适当的响应。本文通过提出一种通用的检测系统来解决这些挑战,旨用于智能城市应用,并在交通和火灾检测场景中进行测试。我们的贡献包括对检测需求的分析以及开发适用于不同应用的系统,推动了智能城市中自动化监视的发展。

关键词

引用

@article{arxiv.2411.15773,
  title  = {Context-Aware Detection of Mixed Critical Events using Video Classification},
  author = {Filza Akhlaq and Alina Arshad and Muhammad Yehya Hayati and Jawwad A. Shamsi and Muhammad Burhan Khan},
  journal= {arXiv preprint arXiv:2411.15773},
  year   = {2025}
}

备注

The results in this paper are old and outdated, as we are working more on the new updated data the results in this manuscript stand invalid. We will update the result on valid data and upload that one here. Thank you for your understanding