Using transfer learning for smart building management system
Open Access
- 7 December 2019
- journal article
- research article
- Published by Springer Nature in Journal of Big Data
- Vol. 6 (1) , 1-12
Abstract
In building management, energy optimization is one of the main concern that needs to be automated. For automation, an intelligent system needs to be developed. However, an intelligent system needs to be trained in a large dataset before it can be used reliably. In this paper, we present a transfer learning scheme to develop an intelligent system for smart building management system. Specifically, the intelligent system is able to count human inside a room, which can be utilized to adaptively adjust energy usage in a room. The transfer learning scheme employs a deep learning model that is pretrained on ImageNet dataset. To enable the human counting capability, the model is trained on a dataset specifically collected for human counting case.Keywords
Funding Information
- Directorate General of Research and Development Strengthening, Indonesian Ministry of Research, Technology, and Higher Education (23/AKM/MONOPNT/2019)
This publication has 35 references indexed in Scilit:
- Faster R-CNN: Towards Real-Time Object Detection with Region Proposal NetworksIEEE Transactions on Pattern Analysis and Machine Intelligence, 2016
- Single-Image Crowd Counting via Multi-Column Convolutional Neural Network2016
- You Only Look Once: Unified, Real-Time Object Detection2016
- Going deeper with convolutions2015
- ImageNet Large Scale Visual Recognition ChallengeInternational Journal of Computer Vision, 2015
- Caffe2014
- Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation2014
- Multi-source Multi-scale Counting in Extremely Dense Crowd Images2013
- The application and implementation research of smart city in China2012
- Backpropagation Applied to Handwritten Zip Code RecognitionNeural Computation, 1989