Hybrid pre trained model based feature extraction for enhanced indoor scene classification in federated learning environments

Abstract Classification of indoor scenes is a crucial task of computer vision. It has widespread applications like smart homes, smart cities, robotics, etc. Primitive classification methods like Support Vector Machines (SVM) and K-Nearest Neighbors (KNN), provide a compromised performance with compl...

Full description

Saved in:
Bibliographic Details
Main Authors: Monica Dutta, Deepali Gupta, Vikas Khullar, Sapna Juneja, Roobaea Alroobaea, Pooja Sapra
Format: Article
Language:English
Published: Nature Portfolio 2025-08-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-025-16673-3
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849226422231498752
author Monica Dutta
Deepali Gupta
Vikas Khullar
Sapna Juneja
Roobaea Alroobaea
Pooja Sapra
author_facet Monica Dutta
Deepali Gupta
Vikas Khullar
Sapna Juneja
Roobaea Alroobaea
Pooja Sapra
author_sort Monica Dutta
collection DOAJ
description Abstract Classification of indoor scenes is a crucial task of computer vision. It has widespread applications like smart homes, smart cities, robotics, etc. Primitive classification methods like Support Vector Machines (SVM) and K-Nearest Neighbors (KNN), provide a compromised performance with complex indoor environments due to light variations, intra-class similarities, and occlusions. Deep Learning (DL) models, especially Convolutional Neural Networks (CNNs), have improved classification accuracy significantly by extracting the image features. This study proposes and implements a novel MultiData model, which integrates DL with Linear Discriminant Analysis (LDA) and Federated Learning (FL) to enhance classification performance while preserving data privacy. Comparative performance analysis of the MultiData model with VGG16, VGG19, and ResNet152 shows that MultiData achieves near-perfect accuracy (99.99%) and minimal validation loss (0%). The performance of the proposed model was further compared with FL-based training across four clients using IID and non-IID datasets. The results further confirm the model’s robustness, achieving 100% accuracy in training and over 95% in validation times respectively. This research is beneficial for stakeholders in healthcare, smart infrastructure, surveillance, and other IoT-based automation, aligning with SDG 9 (Industry, Innovation, and Infrastructure), SDG 11 (Sustainable Cities and Communities), and SDG 12 (Responsible Consumption and Production). By enabling efficient and privacy-preserving scene classification, this work contributes significantly to the development of smart and sustainable environments enhancing AI-based decision-making across various sectors.
format Article
id doaj-art-5659af0f4a1e47c6987b0af72c6a67ef
institution Kabale University
issn 2045-2322
language English
publishDate 2025-08-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-5659af0f4a1e47c6987b0af72c6a67ef2025-08-24T11:21:46ZengNature PortfolioScientific Reports2045-23222025-08-0115111910.1038/s41598-025-16673-3Hybrid pre trained model based feature extraction for enhanced indoor scene classification in federated learning environmentsMonica Dutta0Deepali Gupta1Vikas Khullar2Sapna Juneja3Roobaea Alroobaea4Pooja Sapra5Department of Computer Engineering & Applications, Institute of Engineering & Technology, GLA UniversityChitkara University Institute of Engineering and Technology, Chitkara UniversityChitkara University Institute of Engineering and Technology, Chitkara UniversityKIET Group of InstitutionsDepartment of Computer Science, College of Computers and Information Technology, Taif UniversityParul Institute of Engineering and Technology, Faculty of Engineering and Technology, Parul UniversityAbstract Classification of indoor scenes is a crucial task of computer vision. It has widespread applications like smart homes, smart cities, robotics, etc. Primitive classification methods like Support Vector Machines (SVM) and K-Nearest Neighbors (KNN), provide a compromised performance with complex indoor environments due to light variations, intra-class similarities, and occlusions. Deep Learning (DL) models, especially Convolutional Neural Networks (CNNs), have improved classification accuracy significantly by extracting the image features. This study proposes and implements a novel MultiData model, which integrates DL with Linear Discriminant Analysis (LDA) and Federated Learning (FL) to enhance classification performance while preserving data privacy. Comparative performance analysis of the MultiData model with VGG16, VGG19, and ResNet152 shows that MultiData achieves near-perfect accuracy (99.99%) and minimal validation loss (0%). The performance of the proposed model was further compared with FL-based training across four clients using IID and non-IID datasets. The results further confirm the model’s robustness, achieving 100% accuracy in training and over 95% in validation times respectively. This research is beneficial for stakeholders in healthcare, smart infrastructure, surveillance, and other IoT-based automation, aligning with SDG 9 (Industry, Innovation, and Infrastructure), SDG 11 (Sustainable Cities and Communities), and SDG 12 (Responsible Consumption and Production). By enabling efficient and privacy-preserving scene classification, this work contributes significantly to the development of smart and sustainable environments enhancing AI-based decision-making across various sectors.https://doi.org/10.1038/s41598-025-16673-3Indoor scene classificationComputer visionDeep learningFederated learning
spellingShingle Monica Dutta
Deepali Gupta
Vikas Khullar
Sapna Juneja
Roobaea Alroobaea
Pooja Sapra
Hybrid pre trained model based feature extraction for enhanced indoor scene classification in federated learning environments
Scientific Reports
Indoor scene classification
Computer vision
Deep learning
Federated learning
title Hybrid pre trained model based feature extraction for enhanced indoor scene classification in federated learning environments
title_full Hybrid pre trained model based feature extraction for enhanced indoor scene classification in federated learning environments
title_fullStr Hybrid pre trained model based feature extraction for enhanced indoor scene classification in federated learning environments
title_full_unstemmed Hybrid pre trained model based feature extraction for enhanced indoor scene classification in federated learning environments
title_short Hybrid pre trained model based feature extraction for enhanced indoor scene classification in federated learning environments
title_sort hybrid pre trained model based feature extraction for enhanced indoor scene classification in federated learning environments
topic Indoor scene classification
Computer vision
Deep learning
Federated learning
url https://doi.org/10.1038/s41598-025-16673-3
work_keys_str_mv AT monicadutta hybridpretrainedmodelbasedfeatureextractionforenhancedindoorsceneclassificationinfederatedlearningenvironments
AT deepaligupta hybridpretrainedmodelbasedfeatureextractionforenhancedindoorsceneclassificationinfederatedlearningenvironments
AT vikaskhullar hybridpretrainedmodelbasedfeatureextractionforenhancedindoorsceneclassificationinfederatedlearningenvironments
AT sapnajuneja hybridpretrainedmodelbasedfeatureextractionforenhancedindoorsceneclassificationinfederatedlearningenvironments
AT roobaeaalroobaea hybridpretrainedmodelbasedfeatureextractionforenhancedindoorsceneclassificationinfederatedlearningenvironments
AT poojasapra hybridpretrainedmodelbasedfeatureextractionforenhancedindoorsceneclassificationinfederatedlearningenvironments