Scholarlink

For effective traffic management and road safety, traffic sign recognition is essential. Here, we provide a novel method for traffic sign recognition that makes use of the potent EfficientNetB0-CNN transfer learning model. This study fine-tunes EfficientNetB0, which is well-known for its remarkable performance in picture classification tasks, with a tailored Convolutional Neural Network (CNN) for traffic sign identification. The training dataset for the model includes a wide range of traffic signs that are subjected to different lighting situations, weather fluctuations, and road surroundings. The model can use pre-trained weights thanks to the transfer learning process, which speeds up convergence and boosts performance. The model is extensively tested and its hyper parameters are fine-tuned in order to maximize its precision and effectiveness. The outcomes reveal that the suggested EfficientNetB0-CNN model recognizes traffic signs with amazing accuracy and resilience, highlighting its potential for use in autonomous cars and traffic safety systems. The suggested method creates new opportunities for improving technologies for recognizing traffic signs, which will make transportation systems safer and more intelligent.

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

An Effective Netb0-CNN Transfer Learning Model for Traffic Sign Recognition and Safety

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Suraya Mubeen,
  • Attuluri Uday Kiran,
  • P Raveendra Babu,
  • A. Pradeep Kumar,
  • Madhavi Pingili,
  • Bommireddy Prasanthi

摘要

For effective traffic management and road safety, traffic sign recognition is essential. Here, we provide a novel method for traffic sign recognition that makes use of the potent EfficientNetB0-CNN transfer learning model. This study fine-tunes EfficientNetB0, which is well-known for its remarkable performance in picture classification tasks, with a tailored Convolutional Neural Network (CNN) for traffic sign identification. The training dataset for the model includes a wide range of traffic signs that are subjected to different lighting situations, weather fluctuations, and road surroundings. The model can use pre-trained weights thanks to the transfer learning process, which speeds up convergence and boosts performance. The model is extensively tested and its hyper parameters are fine-tuned in order to maximize its precision and effectiveness. The outcomes reveal that the suggested EfficientNetB0-CNN model recognizes traffic signs with amazing accuracy and resilience, highlighting its potential for use in autonomous cars and traffic safety systems. The suggested method creates new opportunities for improving technologies for recognizing traffic signs, which will make transportation systems safer and more intelligent.