Using transfer learning for smart building management system

Pardamean, Bens; Muljo, Hery Harjono; Cenggoro, Tjeng Wawan; Chandra, Bloomest Jansen; Rahutomo, Reza

doi:10.1186/s40537-019-0272-6

Cited by 32 publications

(7 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The Gap Regularizer module can be attached on top of any deep-learning-based people counting models as the backbone network by first removing the original prediction layers. By having the backbone network be pretrained, the whole training process can be viewed as transfer learning, which has been proven to be beneficial in streamline computer vision tasks ( Cenggoro, 2020 ; Girshick et al, 2014 ; Kornblith, Shlens & Le, 2019 ; Pardamean et al, 2018 , 2019 ). Interestingly, with the design of Gap Regularizer, the information from both the people inside and outside of the RoI is able to flow to the backbone network, resulting in a backbone network that can learn to differentiate between the inside and outside people.…”

Section: Proposed Methodsmentioning

confidence: 99%

“…The problem of counting people in an RoI was formally defined by Pardamean et al (2019) in the case of developing a smart building management system. In their study, direct regression was utilized instead of the popular density map regression.…”

Section: Literature Reviewmentioning

confidence: 99%

“…One of the prevailing models is people counting, which is defined as an automatic system that counts the number of people captured in CCTV footage. It has been applied to numerous cases, such as crowd monitoring ( Chan, Liang & Vasconcelos, 2008 ), scene understanding ( Zhang et al, 2015 ; Shao et al, 2015 ; Zhou, Wang & Tang, 2012 ; Marcellino, Cenggoro & Pardamean, 2022 ), and smart buildings ( Pardamean et al, 2021 , 2019 ). Not only for people, automatic counting is also important in other fields such as agriculture ( Alkhudaydi, Zhou & De La lglesia, 2019 ; Tu et al, 2020 ), ecological surveys ( Arteta, Lempitsky & Zisserman, 2016 ), and traffic monitoring ( Zhang et al, 2017 ; Ciampi et al, 2021 ).…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Counting people inside a region-of-interest in CCTV footage with deep learning

Pardamean

Abid

Cenggoro

et al. 2022

PeerJ Computer Science

Self Cite

View full text Add to dashboard Cite

In recent years, the performance of people-counting models has been dramatically increased that they can be implemented in practical cases. However, the current models can only count all of the people captured in the inputted closed circuit television (CCTV) footage. Oftentimes, we only want to count people in a specific Region-of-Interest (RoI) in the footage. Unfortunately, simple approaches such as covering the area outside of the RoI are not applicable without degrading the performance of the models. Therefore, we developed a novel learning strategy that enables a deep-learning-based people counting model to count people only in a certain RoI. In the proposed method, the people counting model has two heads that are attached on top of a crowd counting backbone network. These two heads respectively learn to count people inside the RoI and negate the people count outside the RoI. We named this proposed method Gap Regularizer and tested it on ResNet-50, ResNet-101, CSRNet, and SFCN. The experiment results showed that Gap Regularizer can reduce the mean absolute error (MAE), root mean square error (RMSE), and grid average mean error (GAME) of ResNet-50, which is the smallest CNN model, with the highest reduction of 45.2%, 41.25%, and 46.43%, respectively. On shallow models such as the CSRNet, the regularizer can also drastically increase the SSIM by up to 248.65% in addition to reducing the MAE, RMSE, and GAME. The Gap Regularizer can also improve the performance of SFCN which is a deep CNN model with back-end features by up to 17.22% and 10.54% compared to its standard version. Moreover, the impacts of the Gap Regularizer on these two models are also generally statistically significant (P-value < 0.05) on the MOT17-09, MOT20-02, and RHC datasets. However, it has a limitation in which it is unable to make significant impacts on deep models without back-end features such as the ResNet-101.

show abstract

Section: Proposed Methodsmentioning

confidence: 99%

Section: Literature Reviewmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Counting people inside a region-of-interest in CCTV footage with deep learning

Pardamean

Abid

Cenggoro

et al. 2022

PeerJ Computer Science

Self Cite

View full text Add to dashboard Cite

show abstract

“…In the commercial building, the places used for smart building research consist of hotels, office buildings, and campus/school buildings. In [22], [29], [37], [49] 4 Campus/school building [32], [34], [40], [45], [52], [78] 6…”

Section: Implementation Of Smart Building In the Property Industrymentioning

confidence: 99%

Smart building trend, role, and position: a systematic literature review

Saputra

Ramadhan²

2023

IJRES

View full text Add to dashboard Cite

<p>Property growth, especially in high-rise buildings in developed and developing countries, has experienced a significant increase. The increasingly expensive resources and the development of information technology encourage the growth of smart buildings in almost all countries, especially in developed and developing countries, and become a trend. The Internet of things (IoT) is the main driver in the development of smart buildings. Property businesses are competing to adopt and implement smart buildings on their properties. This is further strengthened by the development of smart cities in almost all cities around the world, one of the criteria is a smart community. The purpose of implementing smart buildings is to make property management effective and efficient. Using the systematic literature review, this paper will discuss what components must be met for a property called a smart building, then what is the role of smart buildings for an area or community, and how the current smart building trend will be in the future.</p>

show abstract

“…In our country, although some manufacturers have developed various application software that meet the OPC technical standards, there is still a significant gap in performance and other aspects compared with foreign countries. However, with the popularization and use of OPC technology, the integration of building automation systems based on OPC technology will become a hot topic in international research [5][6].…”

Section: Introductionmentioning

confidence: 99%

Intelligent OPC Building Information Management System Based on IoT

Wang

Tao

et al. 2021

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

With the rapid development of artificial intelligence, computers and Internet of Things technologies, building intelligence is becoming more and more popular. However, as the various functional subsystems and equipment of intelligent buildings are connected to each other, the traditional management system needs to monitor and modulate more and more programs, which will have a great impact on the interoperability of the subsystems. Therefore, a more complete and powerful management system is needed. OPC provides a unified standard that can effectively solve this problem. Therefore, the purpose of this article is to design an intelligent OPC building information management system based on the Internet of Things. This article first summarizes the development history and status quo of OPC technology, and then extends the design principles of building information management system based on OPC. Based on its design principles, a detailed analysis of the various subsystems of the building information management system, such as fire protection, intrusion prevention, monitoring, access control, and central air conditioning, is carried out. This article systematically explained the application of PID in the building information management system. And use comparative analysis method, observation method and other research forms to conduct experimental research on the intelligent OPC building information management system based on computer Internet of things. The research shows that compared with the traditional building management system, the intelligent OPC-based building management system researched in this paper can transmit information faster and have higher accuracy.

show abstract

Using transfer learning for smart building management system

Cited by 32 publications

References 36 publications

Counting people inside a region-of-interest in CCTV footage with deep learning

Counting people inside a region-of-interest in CCTV footage with deep learning

Smart building trend, role, and position: a systematic literature review

Intelligent OPC Building Information Management System Based on IoT

Contact Info

Product

Resources

About