Scholarlink

Many road accidents happen due to human mistakes, such as driver fatigue, lack of concentration, and distraction. Self-driving technology promises to reduce incidents caused by these factors. It requires the development of various applications, including systems of traffic sign detection. However, there are significant challenges in constructing such systems, especially when detecting minute objects like remote or small traffic signs, which are crucial for safe automated vehicles. Several researchers have tried to solve this problem from different angles. In this work, we review some of these solutions and propose our multi-class model for classifying traffic signs, in which we employ transfer learning with Support Vector Machine SVM and AlexNet. In term of accuracy, the proposed classifier achieves an accuracy of 86%.

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

Machine Learning Approach for the Classification to Overcome Traffic Sign Detection Challenge

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Fatima Qanouni,
  • Hakim El Massari,
  • Noreddine Gherabi,
  • Maria El Badaoui

摘要

Many road accidents happen due to human mistakes, such as driver fatigue, lack of concentration, and distraction. Self-driving technology promises to reduce incidents caused by these factors. It requires the development of various applications, including systems of traffic sign detection. However, there are significant challenges in constructing such systems, especially when detecting minute objects like remote or small traffic signs, which are crucial for safe automated vehicles. Several researchers have tried to solve this problem from different angles. In this work, we review some of these solutions and propose our multi-class model for classifying traffic signs, in which we employ transfer learning with Support Vector Machine SVM and AlexNet. In term of accuracy, the proposed classifier achieves an accuracy of 86%.