Ahmed Mujtaba

Ahmed Mujtaba, Gleb Radchenko, Marc Masana et al.

Driver fatigue is a critical safety concern in advanced driver assistance systems. Driver monitoring models trained off-site on static datasets adapt poorly to real-world conditions, while standard federated learning imposes high communication overhead, assumes homogeneous architectures, and struggles with personalized driver data. We present FedADAS, a federated distillation framework enabling collaborative on-device learning across heterogeneous vehicular networks. FedADAS enables full model heterogeneity by exchanging only soft logits on a shared public dataset, allowing each vehicle to run a customized model tailored to its computational constraints. Additionally, we introduce a yawn recognition pipeline supporting training and inference on edge devices that provides two robust architectures: Performance-Efficient (99.7 MB) achieving 98.3% F1-score with 1.99ms inference time on a Jetson NANO, and a Memory-Efficient (0.6 MB) that trains an epoch in 6.12 minutes on a Jetson AGX Orin. In experiments with up to 115 edge clients, FedADAS significantly outperforms traditional federated learning approaches at higher client participation, achieving up to 9974x reduction in communication cost while maintaining a superior tradeoff between personalization and generalization under extreme data heterogeneity, demonstrating its suitability for real-world deployment. Code is available at https://opensource.silicon-austria.com/mujtabaa/fedadas

Ahmed Mujtaba, Gleb Radchenko, Radu Prodan et al.

Federated distillation has emerged as a promising collaborative machine learning approach, offering enhanced privacy protection and reduced communication compared to traditional federated learning by exchanging model outputs (soft logits) rather than full model parameters. However, existing methods employ complex selective knowledge-sharing strategies that require clients to identify in-distribution proxy data through computationally expensive statistical density ratio estimators. Additionally, server-side filtering of ambiguous knowledge introduces latency to the process. To address these challenges, we propose a robust, resource-efficient EdgeFD method that reduces the complexity of the client-side density ratio estimation and removes the need for server-side filtering. EdgeFD introduces an efficient KMeans-based density ratio estimator for effectively filtering both in-distribution and out-of-distribution proxy data on clients, significantly improving the quality of knowledge sharing. We evaluate EdgeFD across diverse practical scenarios, including strong non-IID, weak non-IID, and IID data distributions on clients, without requiring a pre-trained teacher model on the server for knowledge distillation. Experimental results demonstrate that EdgeFD outperforms state-of-the-art methods, consistently achieving accuracy levels close to IID scenarios even under heterogeneous and challenging conditions. The significantly reduced computational overhead of the KMeans-based estimator is suitable for deployment on resource-constrained edge devices, thereby enhancing the scalability and real-world applicability of federated distillation. The code is available online for reproducibility.

Ahmed Mujtaba, Gleb Radchenko, Marc Masana et al.

Driver fatigue remains a leading cause of road accidents, with 24% of crashes involving drowsy drivers. While yawning serves as an early behavioral indicator of fatigue, existing machine learning approaches face significant challenges due to video-annotated datasets that introduce systematic noise from coarse temporal annotations. We develop a semi-automated labeling pipeline with human-in-the-loop verification, which we apply to YawDD, enabling more accurate model training. 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. The resulting approach deliver up to 59.8 FPS on edge AI hardware (NVIDIA Jetson Nano), confirming that enhanced data quality alone supports on-device yawning monitoring without server-side computation.