Article Per Year (5 Year)
p-Index From 2021 - 2026
0.61
P-Index
This Author published in this journals
All Journal Sinkron : Jurnal dan Penelitian Teknik Informatika Building of Informatics, Technology and Science
Maulana, Isa Iant
Unknown Affiliation
Computer Science & IT

Published : 3 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 2 Documents
Search
Journal : Building of Informatics, Technology and Science

 1 
Comprehensive Benchmark of Yolov11n, SSD MobileNet, CenterFace, Yunet, FastMtCnn, HaarCascade, and LBP for Face Detection in Video Based Driver Drowsiness Go, Agnestia Agustine Djoenaidi; Alzami, Farrikh; Naufal, Muhammad; Azies, Harun Al; Winarno, Sri; Pramunendar, Ricardus Anggi; Megantara, Rama Aria; Maulana, Isa Iant; Arif, Mohammad
Building of Informatics, Technology and Science (BITS) Vol 7 No 3 (2025): December 2025
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v7i3.8678

Abstract

Face detection is a critical foundation of video-based drowsiness monitoring systems because all downstream tasks such as eye-closure estimation, yawning detection, and head movement analysis depend entirely on correctly identifying the face region. Many previous studies rely on detector-generated outputs as ground truth, which can introduce bias and inflate model performance . To avoid this limitation, I manually constructed a ground truth dataset using 1,229 frames extracted from 129 yawning and microsleep videos in the NITYMED dataset. Ten representative frames were sampled from each video using a face-guided extraction script, and all frames were manually annotated in Roboflow following the COCO format to ensure accurate bounding box labeling under varying lighting, head poses, and facial deformation. Using this manually annotated dataset, I conducted a comprehensive benchmark of seven face-detection algorithms: YOLOv11n, SSD MobileNet, CenterFace, YuNet, FastMtCnn, HaarCascade, and LBP. The evaluation focused on localization quality using Intersection over Union (IoU ≥ 0.5) and Dice Similarity, allowing each algorithm’s predicted bounding box to be directly compared against human defined ground truth. The results show that HaarCascade achieved the highest IoU and Dice scores, particularly in frontal and well-lit frames. FastMtCnn also produced strong alignment with a high number of correctly matched frames. CenterFace and SSD MobileNet demonstrated smooth bounding box fitting with competitive Dice scores, while YOLOv11n and YuNet delivered moderate but stable performance across most samples. LBP showed the weakest results, mainly due to its sensitivity to lighting variations and soft-texture regions. Overall, this benchmark provides an unbiased and comprehensive comparison of modern and classical face-detection algorithms for video-based driver-drowsiness applications.
Dampak Penggunaan Data Augmentasi Terhadap Akurasi MobileNetV2 Dalam Deteksi Mikrosleep Berbasis Rasio Aspek Mata Maulana, Isa Iant; Riadi, Muhammad Fatah Abiyyu; Alzami, Farrikh; Naufal, Muhammad; Azies, Harun Al; Pramunendar, Ricardus Anggi; Basuki, Ruri Suko
Building of Informatics, Technology and Science (BITS) Vol 7 No 3 (2025): December 2025
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v7i3.8719

Abstract

Detecting microsleep is important in preventing accidents caused by decreased alertness, especially in activities that require high concentration such as driving. This study aims to develop an image-based microsleep detection model using the MediaPipe FaceMesh. The EAR value is only used for the tagging process that forms the basis for dataset creation. The main problem investigated is how to produce a classification model that can accurately distinguish between normal eye conditions and microsleep conditions using image data taken from eye area snippets. To address this issue, this study applies a series of stages, starting from dataset formation, initial processing in the form of image size adjustment, normalization, and quality improvement through data augmentation, to model training using the MobileNetV2 architecture with transfer learning and fine-tuning techniques. The results of the experiment show that the use of data augmentation strategies has a significant effect on improving model performance, with the best configuration producing a test accuracy of 87.54 percent, with other high performance metrics, namely Precision of 88.64 percent, Recall (Sensitivity) of 87.14 percent, and F1-Score of 87.34 percent. These findings prove that an eye area image-based approach combined with a convolutional neural network model is capable of providing promising performance in detecting microsleep conditions. These findings prove that an approach based on eye area images combined with a convolutional neural network model can deliver promising performance in detecting microsleep. This research is expected to form the basis for the development of a more effective microsleep detection system that can be implemented in real world environments.
Co-Authors Alzami, Farrikh Dewi Agustini Santoso Go, Agnestia Agustine Djoenaidi Harun Al Azies Ifan Rizqa Megantara, Rama Aria Mohammad Arif Muhammad Naufal, Muhammad Nugraini, Siti Hadiati Riadi, Muhammad Fatah Abiyyu Ricardus Anggi Pramunendar Ruri Suko Basuki Sofiani, Hilda Ayu Sri Winarno