Scholarlink

<p>With the rapid development of artificial intelligence, image recognition based on deep learning (DL) is changing traditional industries like the hotel industry. The purpose of this study is to improve the intelligence level of hotel operation and management by solving the problems of high dependence, slow response and data fragmentation of hotel operation and management personnel. In this study, a visual management strategy based on DL is proposed; this is tailored for various hotel scenes such as front desk, guest room and kitchen by combining with image recognition. The proposed methods include evaluating the cleanliness of guest rooms, identifying customer behaviors, and detecting abnormal events early; it achieves an accuracy of 0.90 and a false positive rate of only 0.08 on the RoomNet dataset by implementing an optimized model fusing a convolutional neural network and a You Only Look Once (YOLO) structure. The model is superior to YOLOv8 and ConvNeXt in recognition performance and operation efficiency. It shows superior management assistance abilities. This visual management platform can present the image recognition results and automatic task scheduling in real time. Meanwhile, it can improve the human replacement and task automation rates to 80.29% and 82.41% in the guest room scenario. These findings provide a theoretical basis and practical path for the integrated application of hotel intelligent operation and image recognition.</p>

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

Deep learning-based image recognition for visual management and intelligent decision support in hotel operations

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月10日

查看全文
  • Jun Zhang,
  • Guohao Zhong,
  • Yunsheng Xia

摘要

With the rapid development of artificial intelligence, image recognition based on deep learning (DL) is changing traditional industries like the hotel industry. The purpose of this study is to improve the intelligence level of hotel operation and management by solving the problems of high dependence, slow response and data fragmentation of hotel operation and management personnel. In this study, a visual management strategy based on DL is proposed; this is tailored for various hotel scenes such as front desk, guest room and kitchen by combining with image recognition. The proposed methods include evaluating the cleanliness of guest rooms, identifying customer behaviors, and detecting abnormal events early; it achieves an accuracy of 0.90 and a false positive rate of only 0.08 on the RoomNet dataset by implementing an optimized model fusing a convolutional neural network and a You Only Look Once (YOLO) structure. The model is superior to YOLOv8 and ConvNeXt in recognition performance and operation efficiency. It shows superior management assistance abilities. This visual management platform can present the image recognition results and automatic task scheduling in real time. Meanwhile, it can improve the human replacement and task automation rates to 80.29% and 82.41% in the guest room scenario. These findings provide a theoretical basis and practical path for the integrated application of hotel intelligent operation and image recognition.