Article Per Year (5 Year)
p-Index From 2019 - 2024
0.23
P-Index
This Author published in this journals
All Journal SINTECH (Science and Information Technology) Journal
Ryandra Anditto
President University
Computer Science & IT

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
SECURITY MONITORING USING IMPROVED MOBILENET V2 WITH FINE-TUNING TO PREVENT THEFT IN RESIDENTIAL AREAS DURING THE COVID-19 PANDEMIC Ryandra Anditto; Rusdianto Roestam
SINTECH (Science and Information Technology) Journal Vol. 5 No. 1 (2022): SINTECH Journal Edition April 2022
Publisher : Prahasta Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31598/sintechjournal.v5i1.1023

Abstract

In this journal, authors improved a deep learning, it was MobileNet V2 which can learn with higher speed and accuracy using the ImageNet dataset. The basic architecture of MobileNet V2 was modified using a fine-tuning technique. It did not change the entire weight of the default deep learning model, there was freezing of some layers. The modifications involved by changing the parameters on the layer. The improved model would be trained using the ImageNet dataset based on the security monitoring context. The dataset class objects selected for training were objects that usually used by suspicious people with the aim of committing crime of theft. The results of training using an improved model could increase accuracy up to 71% with a difference of 3% from the training results of the default model of MobileNet V2. Since MobileNet V2 was a lightweight deep neural network model that had few parameters compared to other neural network model parameters, this modified model could be implemented on devices with low specifications, such as mobile devices or raspberry pi in the form of real-time applications.
Co-Authors Rusdianto Roestam