Advancing Cognitive Load Detection in Simulated Driving Scenarios Through Deep Learning and fNIRS Data
DOI: 10.3390/s25164921
archive: archived pipeline: cataloged verified
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This study addresses the critical safety challenge of driver cognitive overload in conditionally automated vehicles, where human factors remain a primary cause of accidents. The research aims to detect varying levels of cognitive load in real-time using Functional Near-Infrared Spectroscopy (fNIRS) and deep learning within a high-fidelity driving simulator. The motivation stems from the need for robust, scalable monitoring systems that can handle the complex, non-linear patterns of brain activity associated with multitasking under stressful environmental conditions, such as night-time driving and heavy rainfall. The experimental design involved 38 healthy participants with valid driver’s licenses who performed a dual-task paradigm: driving a simulated Toyota Fortuner SUV in challenging weather conditions while simultaneously completing an auditory-modified n-back task. This secondary task induced three graded levels of cognitive workload: 0-back (baseline), 1-back, and 2-back (highest load). fNIRS data capturing hemodynamic changes in the prefrontal cortex was recorded using a high-density wearable system. The raw signals were pre-processed and segmented into temporal windows of 10, 20, and 30 seconds using both overlapping and non-overlapping strategies. These segments were then classified using EEGNet, a compact convolutional neural network originally designed for EEG data, which was adapted here to extract spatiotemporal features from fNIRS signals. The model was trained using the Adam optimizer over 200 epochs with categorical cross-entropy loss. The results demonstrated that classification performance was significantly influenced by the learning rate and the windowing method employed. A learning rate of 0.001 yielded the highest accuracy. Specifically, the model achieved 100% accuracy when using overlapping windows and 97% accuracy with non-overlapping windows. These findings indicate that the sliding window approach effectively captures short-term fluctuations and transient changes in brain activation, enhancing the model's ability to distinguish between different cognitive load levels. The study confirms that incremental increases in cognitive demand, particularly during the 2-back task, correspond to detectable changes in cerebral blood oxygenation patterns that can be accurately classified by deep learning models. The significance of this work lies in its demonstration of the potential for combining fNIRS with deep learning for real-time cognitive load monitoring in realistic driving scenarios. By achieving near-perfect classification accuracy, the study highlights the importance of temporal modeling in physiological signal analysis and suggests that such systems can generalize across participants better than traditional statistical methods. These findings support the development of adaptive driver-assistance systems capable of detecting mental fatigue and distraction, thereby improving safety in partially automated vehicles.
Key finding
A deep learning model using fNIRS data achieved 100% accuracy in classifying cognitive load levels when using overlapping temporal windows and a learning rate of 0.001.
Methodology
simulator
Sample size: 38
Provenance
The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed. Discovered via author_sweep_intake on 2026-05-28.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | author_sweep | — | — | 2 | 2026-05-28 |
| archive | success | canonical_url | — | — | 11 | 2026-06-06 |
| extract | success | cached | — | — | 3 | 2026-06-10 |
| clean | success | clean | — | — | 1 | 2026-06-04 |
| chunk | success | chunk | — | — | 1 | 2026-06-04 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-04 |
| enrich | success | — | — | — | 1 | 2026-05-28 |
| promote | success | — | — | — | 1 | 2026-06-04 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 2 | 2026-06-10 |
| tag | success | vector_similarity | — | — | 15 | 2026-06-11 |
| verify | success | — | — | — | 2 | 2026-06-10 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.
Information type
What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).
- Empirical Findings: physiological data
- Methodological Resource: tool software
- Theoretical Contribution: computational model