Scholarlink

<p>Occupancy detection is highly beneficial in smart buildings and smart homes for automation and optimization of power consumption. This paper presents a Neural Network model utilizing various environmental sensors to predict occupancy. While different statistical machine learning models for occupancy detection are available in the literature, they have not been thoroughly analyzed for deployment on edge computing platforms. Edge devices typically have limited computational resources. Therefore, it is crucial to analyze the performance of machine learning models on edge platforms before deployment. In this study, the effectiveness of the proposed approach is tested on various 32-bit microcontrollers. The results demonstrate that the lowest inference time is achieved using the STM32H735G microcontroller, which operates at 550&#xa0;MHz.</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

Benchmarking and Performance Analysis of a Deep Learning Model on Edge Computing Platforms for Occupancy Detection in Smart Buildings

  • 期刊论文

  • 发表日期：

    2025年9月22日

查看全文
  • Ritesh Kumar Sharma,
  • Gaurav Verma

摘要

Occupancy detection is highly beneficial in smart buildings and smart homes for automation and optimization of power consumption. This paper presents a Neural Network model utilizing various environmental sensors to predict occupancy. While different statistical machine learning models for occupancy detection are available in the literature, they have not been thoroughly analyzed for deployment on edge computing platforms. Edge devices typically have limited computational resources. Therefore, it is crucial to analyze the performance of machine learning models on edge platforms before deployment. In this study, the effectiveness of the proposed approach is tested on various 32-bit microcontrollers. The results demonstrate that the lowest inference time is achieved using the STM32H735G microcontroller, which operates at 550 MHz.