SynthSecureNet: An Improved Deep Learning Architecture with Application to Intelligent Violence Detection
We present a new deep learning architecture, named SynthSecureNet, which hybridizes two popular architectures: MobileNetV2 and ResNetV2. The latter have been shown to be promising in violence detection. The aim of our architecture is to harness the combined strengths of the two known methods for imp...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-01-01
|
| Series: | Algorithms |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1999-4893/18/1/39 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832589445734334464 |
|---|---|
| author | Ntandoyenkosi Zungu Peter Olukanmi Pitshou Bokoro |
| author_facet | Ntandoyenkosi Zungu Peter Olukanmi Pitshou Bokoro |
| author_sort | Ntandoyenkosi Zungu |
| collection | DOAJ |
| description | We present a new deep learning architecture, named SynthSecureNet, which hybridizes two popular architectures: MobileNetV2 and ResNetV2. The latter have been shown to be promising in violence detection. The aim of our architecture is to harness the combined strengths of the two known methods for improved accuracy. First, we leverage the pre-trained weights of MobileNetV2 and ResNet50V2 to initialize the network. Next, we fine-tune the network by training it on a dataset of labeled surveillance videos, with a focus on optimizing the fusion process between the two architectures. Experimental results demonstrate a significant improvement in accuracy compared with individual models. MobileNetV2 achieves an accuracy of 90%, while ResNet50V2 achieves a 94% accuracy in violence detection tasks. SynthSecureNet achieves an accuracy of 99.22%, surpassing the performance of individual models. The integration of MobileNetV2 and ResNet50V2 in SynthSecureNet offers a comprehensive solution that addresses the limitations of the existing architectures, paving the way for more effective surveillance and crime prevention strategies. |
| format | Article |
| id | doaj-art-fec9de145b1c40dcadacc7f82eb863f9 |
| institution | Kabale University |
| issn | 1999-4893 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Algorithms |
| spelling | doaj-art-fec9de145b1c40dcadacc7f82eb863f92025-01-24T13:17:34ZengMDPI AGAlgorithms1999-48932025-01-011813910.3390/a18010039SynthSecureNet: An Improved Deep Learning Architecture with Application to Intelligent Violence DetectionNtandoyenkosi Zungu0Peter Olukanmi1Pitshou Bokoro2Department of Electrical and Electronic Engineering Technology, Faculty of Engineering and the Built Environment, University of Johannesburg, Johannesburg 2092, South AfricaDepartment of Electrical and Electronic Engineering Technology, Faculty of Engineering and the Built Environment, University of Johannesburg, Johannesburg 2092, South AfricaDepartment of Electrical and Electronic Engineering Technology, Faculty of Engineering and the Built Environment, University of Johannesburg, Johannesburg 2092, South AfricaWe present a new deep learning architecture, named SynthSecureNet, which hybridizes two popular architectures: MobileNetV2 and ResNetV2. The latter have been shown to be promising in violence detection. The aim of our architecture is to harness the combined strengths of the two known methods for improved accuracy. First, we leverage the pre-trained weights of MobileNetV2 and ResNet50V2 to initialize the network. Next, we fine-tune the network by training it on a dataset of labeled surveillance videos, with a focus on optimizing the fusion process between the two architectures. Experimental results demonstrate a significant improvement in accuracy compared with individual models. MobileNetV2 achieves an accuracy of 90%, while ResNet50V2 achieves a 94% accuracy in violence detection tasks. SynthSecureNet achieves an accuracy of 99.22%, surpassing the performance of individual models. The integration of MobileNetV2 and ResNet50V2 in SynthSecureNet offers a comprehensive solution that addresses the limitations of the existing architectures, paving the way for more effective surveillance and crime prevention strategies.https://www.mdpi.com/1999-4893/18/1/39ensemble modelhybrid modelSynthSecureNet2D CNNdeep transfer leaningMobileNetV2 |
| spellingShingle | Ntandoyenkosi Zungu Peter Olukanmi Pitshou Bokoro SynthSecureNet: An Improved Deep Learning Architecture with Application to Intelligent Violence Detection Algorithms ensemble model hybrid model SynthSecureNet 2D CNN deep transfer leaning MobileNetV2 |
| title | SynthSecureNet: An Improved Deep Learning Architecture with Application to Intelligent Violence Detection |
| title_full | SynthSecureNet: An Improved Deep Learning Architecture with Application to Intelligent Violence Detection |
| title_fullStr | SynthSecureNet: An Improved Deep Learning Architecture with Application to Intelligent Violence Detection |
| title_full_unstemmed | SynthSecureNet: An Improved Deep Learning Architecture with Application to Intelligent Violence Detection |
| title_short | SynthSecureNet: An Improved Deep Learning Architecture with Application to Intelligent Violence Detection |
| title_sort | synthsecurenet an improved deep learning architecture with application to intelligent violence detection |
| topic | ensemble model hybrid model SynthSecureNet 2D CNN deep transfer leaning MobileNetV2 |
| url | https://www.mdpi.com/1999-4893/18/1/39 |
| work_keys_str_mv | AT ntandoyenkosizungu synthsecurenetanimproveddeeplearningarchitecturewithapplicationtointelligentviolencedetection AT peterolukanmi synthsecurenetanimproveddeeplearningarchitecturewithapplicationtointelligentviolencedetection AT pitshoubokoro synthsecurenetanimproveddeeplearningarchitecturewithapplicationtointelligentviolencedetection |