Scholarlink

This study proposes a novel approach for monitoring beehive health through audio analysis. By leveraging Mel-Frequency Cepstral Coefficients (MFCC) as feature descriptors and a compact One-Dimensional Convolutional Neural Network (1DCNN) as the classification model, we achieve a significant improvement in accuracy, reaching 90.40% on a real-world dataset. This surpasses state-of-the-art methods, demonstrating the effectiveness of our approach in distinguishing different beehive states. The findings highlight the potential of audio-based techniques for developing robust bee colony health monitoring systems.

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

Improving Bee States Classification Using MFCC and 1DCNN

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Thi-Thu-Hong Phan,
  • Quoc-Trinh Vo,
  • Cao Vu Bui,
  • Luong Vuong Nguyen,
  • Dinh Nam Vo,
  • Thi Thao Ha

摘要

This study proposes a novel approach for monitoring beehive health through audio analysis. By leveraging Mel-Frequency Cepstral Coefficients (MFCC) as feature descriptors and a compact One-Dimensional Convolutional Neural Network (1DCNN) as the classification model, we achieve a significant improvement in accuracy, reaching 90.40% on a real-world dataset. This surpasses state-of-the-art methods, demonstrating the effectiveness of our approach in distinguishing different beehive states. The findings highlight the potential of audio-based techniques for developing robust bee colony health monitoring systems.