Article Per Year (5 Year)
p-Index From 2021 - 2026
2.233
P-Index
This Author published in this journals
All Journal Sosiohumaniora Sinkron : Jurnal dan Penelitian Teknik Informatika RABIT: Jurnal Teknologi dan Sistem Informasi Univrab InComTech: Jurnal Telekomunikasi dan Komputer Jurnal Sistem Cerdas JSAI (Journal Scientific and Applied Informatics) Jurnal Abdi Masyarakat (JAM) Jurnal Sistem Informasi dan Informatika (SIMIKA) Literatify : Trends in Library Development Jevief : Jurnal Vokasi
Ida Nurhaida
Unknown Affiliation
Arts Humanities Automotive Engineering Civil Engineering, Building, Construction & Architecture Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Economics, Econometrics & Finance Education Electrical & Electronics Engineering Engineering Industrial & Manufacturing Engineering Law, Crime, Criminology & Criminal Justice Library & Information Science Social Sciences Other

Published : 15 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 15 Documents
Search

 1   2 
AI-BASED FACIAL DE-IDENTIFICATION FOR CHILDREN'S DIGITAL PRIVACY Komang Putra Satria Negara; Ida Nurhaida
Jurnal Sistem Informasi dan Informatika (Simika) Vol. 8 No. 2 (2025): Jurnal Sistem Informasi dan Informatika (Simika)
Publisher : Program Studi Sistem Informasi, Universitas Banten Jaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47080/simika.v8i2.3848

Abstract

Social media has become an inseparable part of daily life in today's digital era. Many parents frequently share photos of their children online, exposing them to risks related to privacy and security. This research addresses such issues by developing an Android-based facial de-identification application that utilises the YOLOv8 algorithm to protect minors' privacy. The methodology involves several stages: data collection, pre-processing, model training, and application development. The dataset includes over 2,889 images of children, which were augmented to enhance its size and diversity. YOLOv8, a state-of-the-art object detection algorithm, was trained with these images to achieve high precision and recall in identifying children's faces. The developed application integrates the YOLOv8 model within a user-friendly interface built with Flutter. Results indicate that YOLOv8 effectively detects children's faces with a high precision of 95%, accuracy of 88%, recall of 92%, and mAP50 of 0.977. While the model demonstrates strong performance on training data, there is room for improvement on unseen data. By leveraging YOLOv8 and providing an accessible mobile application, the work allows parents to protect their children's identities online. The application mitigates risks of unauthorised use and exploitation of children's images by enabling facial de-identification, thus promoting safer online practices for families.
FACE DETECTION AND ANTI-SPOOFING ON DESKTOP APPLICATIONS USING YOU ONLY LOOK ONCE Fairo Mahaputranda Faisal; Ida Nurhaida
Jurnal Sistem Informasi dan Informatika (Simika) Vol. 8 No. 2 (2025): Jurnal Sistem Informasi dan Informatika (Simika)
Publisher : Program Studi Sistem Informasi, Universitas Banten Jaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47080/simika.v8i2.3849

Abstract

In the digital era, facial recognition systems have become increasingly vulnerable to spoofing attacks, as demonstrated by cases of identity theft using photos or smartphone screens. This study develops a real-time face liveness detection system using YOLOv8 to address these vulnerabilities. Under controlled laboratory conditions, the system achieved exceptional performance metrics: accuracy of 1.0, precision of 1.0, and recall of 1.0, with a mean Average Precision (mAP) of 0.96. However, this study reveals critical insights about the challenges of real-world deployment, including significant performance degradation under poor lighting conditions where genuine faces were misclassified as spoofed images. Compared to existing methods such as Attention-Based Two-Stream CNN (accuracy: 0.91) and Deep Spatial Gradient approaches (accuracy: 0.90-0.92), our system demonstrates superior performance in controlled environments but highlights the persistent challenge of environmental variability in practical applications. These findings emphasize the need for robust preprocessing techniques and diverse training datasets to bridge the gap between laboratory performance and real-world reliability. The study contributes to understanding the limitations of current face anti-spoofing technologies and provides a foundation for developing more robust systems suitable for practical deployment.
AI Model for Detecting Depression Based on Sleep Pattern Analysis Using Sequence Models Gregorius Rizcy Orlando Pradana; Ida Nurhaida
Jurnal Sistem Cerdas Vol. 9 No. 1 (2026)
Publisher : APIC

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37396/jsc.v9i1.651

Abstract

Depression is generally diagnosed through subjective clinical assessments, so objective biomarkers such as sleep patterns are needed. Unfortunately, conventional machine learning methods often ignore the temporal dynamics of sleep. This study aims to evaluate four Sequence Models architectures (LSTM, Bi-LSTM, GRU, Bi-GRU) to detect indications of depression from 7 days of sequential sleep data. The methodology processes data from 5,782 subjects using six physiological features (oxygen saturation, sleep efficiency, spindle microarchitecture) converted into a 3D matrix. Evaluation uses Precision, Recall, F1-Score, and ROC-AUC metrics to handle imbalanced data. The results prove that the Bidirectional model is more robust in capturing the temporal context holistically. Bi-GRU achieved the highest ROC-AUC score (0.9909), while Bi-LSTM produced the best F1-Score (0.85) and Recall (0.82). The standard GRU was validated as the most computationally efficient model (5 seconds/epoch). Explainable AI analysis confirmed that fast spindle percentage, REM duration, and spindle density are the strongest predictors of affective dysfunction. In conclusion, the Bidirectional architecture has proven reliable in identifying sleep anomalies, providing a solid foundation for real-time IoMT-based psychiatric screening systems.
SMART ATTENDANCE SYSTEM BERBASIS WEB REAL-TIME MENGGUNAKAN FACENET Anggi Saputri; Ida Nurhaida; Revaldo Ilfestra Metzi Zen
Rabit : Jurnal Teknologi dan Sistem Informasi Univrab Vol 11 No 1 (2026): Januari
Publisher : LPPM Universitas Abdurrab

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.36341/rabit.v11i1.7041

Abstract

Manual attendance recording often leads to inefficiency and potential data manipulation. To address these issues, this study develops a web-based attendance system with real-time facial recognition using FaceNet. The model is designed by utilizing Dlib for face detection, FaceNet for facial feature embedding generation, and Support Vector Machine for identity classification. The system is implemented using Flask as the backend, Bootstrap for a responsive user interface, and SQLite as a lightweight database. The research dataset consists of 50 individuals and is used to compare the performance of three feature extraction models, namely FaceNet, MobileNet, and VGG-16. The evaluation results indicate that FaceNet achieves the best performance, with a training accuracy of 99.92%, a testing accuracy of 99.82%, an F1-score of 0.9983, and a training time of 9.74 seconds. MobileNet also demonstrates strong performance, achieving a training accuracy of 99.88%, a testing accuracy of 99.73%, and an F1-score of 0.9973. Meanwhile, VGG-16 shows relatively lower performance, with a training accuracy of 99.71%, a testing accuracy of 99.51%, an F1-score of 0.9951, and a training time of 24.61 seconds. These findings indicate that FaceNet is more effective and efficient in extracting facial features. Therefore, the developed system has the potential to replace conventional attendance methods that are prone to errors while supporting the implementation of a smart campus.
IndoBERT-Based Sentiment Analysis of Indonesian Social Media Discourse on AI-Generated Images Halvino Iqbal Nataprawira; Ida Nurhaida
Sinkron : jurnal dan penelitian teknik informatika Vol. 10 No. 3 (2026): Article Research July 2026
Publisher : Politeknik Ganesha Medan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33395/sinkron.v10i3.16242

Abstract

The rapid emergence of generative artificial intelligence has disrupted creative ecosystems, prompting widespread discourse across Indonesian social media. However, the exact sentiment structure of this public reaction remains empirically unmapped due to the contextual complexities of informal language. The objective of this research is to evaluate the efficacy of contextual language models by fine-tuning IndoBERT and benchmarking it against classical machine learning classifiers—including Complement Naive Bayes, Logistic Regression, and Support Vector Machine—for classifying social media sentiment. A multi-platform dataset comprising 2,981 Indonesian-language posts from X, Reddit, and YouTube was collected and manually annotated into positive, neutral, and negative classes. To address inherent class imbalance, Synthetic Minority Oversampling Technique was applied to classical models, while class-weighted loss and Masked Language Modeling augmentation were utilized for IndoBERT. Performance was evaluated using macro-averaged F1-score across five repeated stratified random splits. IndoBERT achieved a mean macro-F1 of 0.7131 ± 0.0180, outperforming the best classical baseline by approximately 0.12, demonstrating a pronounced advantage in resolving ambiguous neutral discourse. Negative sentiment heavily dominated the corpus at 61.8%, reflecting a prevailing critical stance toward AI-generated imagery concerning ethical and copyright issues. Furthermore, evaluation variance across random seeds exceeded variance from augmentation strategies, indicating test set composition is a major performance determinant. In conclusion, this study establishes a robust empirical baseline for Indonesian sentiment analysis, proving transformer architectures superior for nuanced public opinion mining.
Co-Authors Andi Windah Andi Windah Anggi Saputri Ayumi, Vina Bangun Suharti Cahyono Eko Sugiarto Devanna Alandra Khansa Fairo Mahaputranda Faisal Gregorius Rizcy Orlando Pradana Halvino Iqbal Nataprawira Komang Putra Satria Negara Mochamad Sobiri Nina Yudha Aryanti Noprisson, Handrie Ranida Pradita Revaldo Ilfestra Metzi Zen Safitri Jaya Wachyu Hari Haji