Saved in:
Bibliographic Details
Main Authors: Mujtaba, Ahmed, Radchenko, Gleb, Masana, Marc, Prodan, Radu
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2512.11446
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909053443112960
author Mujtaba, Ahmed
Radchenko, Gleb
Masana, Marc
Prodan, Radu
author_facet Mujtaba, Ahmed
Radchenko, Gleb
Masana, Marc
Prodan, Radu
contents Driver fatigue remains a leading cause of road accidents, responsible for 24% of crashes. While yawning serves as an early behavioral indicator of fatigue, existing approaches face significant challenges due to the presence of systematic noise in video-annotated datasets arising from coarse temporal annotations. Training robust machine learning (ML) models requires rich supervisory labels that help learn salient features from the training data. Moreover, efficient on-device training and inference of models on edge devices is crucial in driver fatigue detection tasks to enable accurate real-time decisions on vehicles without reliance on cloud infrastructure. To address this issue, we develop a semi-automated labeling pipeline with human-in-the-loop verification to annotate YawDD videos to YawDD+ frame-level annotations, enabling more accurate model training on edge platforms such as NVIDIA Jetson NANO. Training the established MNasNet classifier and YOLOv11 detector architectures on YawDD+ improves frame accuracy by up to 6% and mAP by 5% over video-level supervision, achieving 99.34% classification accuracy and 95.69% detection mAP on Jetson NANO and AGX. Moreover, MNasNet completed the epoch time in just 8.69 min/epoch while delivering up to 115 frames-per-second (FPS) inference time on AGX, confirming that enhanced data quality alone supports on-device driver fatigue monitoring systems without server-side computation. The YawDD+ dataset and trained models are available online.
format Preprint
id arxiv_https___arxiv_org_abs_2512_11446
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle YawDD+: Frame-level Annotations for Accurate Yawn Prediction
Mujtaba, Ahmed
Radchenko, Gleb
Masana, Marc
Prodan, Radu
Computer Vision and Pattern Recognition
Driver fatigue remains a leading cause of road accidents, responsible for 24% of crashes. While yawning serves as an early behavioral indicator of fatigue, existing approaches face significant challenges due to the presence of systematic noise in video-annotated datasets arising from coarse temporal annotations. Training robust machine learning (ML) models requires rich supervisory labels that help learn salient features from the training data. Moreover, efficient on-device training and inference of models on edge devices is crucial in driver fatigue detection tasks to enable accurate real-time decisions on vehicles without reliance on cloud infrastructure. To address this issue, we develop a semi-automated labeling pipeline with human-in-the-loop verification to annotate YawDD videos to YawDD+ frame-level annotations, enabling more accurate model training on edge platforms such as NVIDIA Jetson NANO. Training the established MNasNet classifier and YOLOv11 detector architectures on YawDD+ improves frame accuracy by up to 6% and mAP by 5% over video-level supervision, achieving 99.34% classification accuracy and 95.69% detection mAP on Jetson NANO and AGX. Moreover, MNasNet completed the epoch time in just 8.69 min/epoch while delivering up to 115 frames-per-second (FPS) inference time on AGX, confirming that enhanced data quality alone supports on-device driver fatigue monitoring systems without server-side computation. The YawDD+ dataset and trained models are available online.
title YawDD+: Frame-level Annotations for Accurate Yawn Prediction
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2512.11446