A comparative and critical study of EEGNet for fNIRS-driven cognitive load classification
Authors:
- Mehshan KhanMKMehshan Khan
- Houshyar AsadiHoushyar Asadi
- Li ZhangLi Zhang
- Mohammad Reza Chalak QazaniMohammad Reza Chalak Qazani
- Ghazal BargshadyGBGhazal Bargshady
- Stefanos GkikasSGStefanos Gkikas
- Christian ArzateCAChristian Arzate
- Sam OladazimiSOSam Oladazimi
- Zoran NajdovskiZNZoran Najdovski
- Lei WeiLWLei Wei
- Chee Peng LimCLChee Peng Lim
Preprint
- 6 September 2026
- preprint
- Published by Elsevier in SSRN Electronic Journal
Abstract
Accurately classifying cognitive load from functional near-infrared spectroscopy (fNIRS) signals remains a significant challenge due to temporal variability, inter-subject differences, and sensitivity to preprocessing choices. This study provides a comprehensive evaluation of EEGNet for fNIRS-based cognitive load classification by systematically examining the effects of temporal segmentation strategies (overlapping vs. non-overlapping), window lengths (10s, 20s, 30s), feature extraction methods (Analysis of Variance (ANOVA), Principal Component Analysis (PCA), Fast Independent Component Analysis (FastICA)), learning rate configurations (fixed and adaptive), and evaluation protocols (random split vs. subject-independent (SI)). Results from random-split experiments show that overlapping segmentation, combined with smaller fixed learning rates (0.01-0.001), yields the highest accuracies, due to temporal redundancy and dense sampling of hemodynamic transitions. However, SI evaluation reveals a substantial drop in accuracy, demonstrating limited generalization to unseen participants. Under SI evaluation, non-overlapping segmentation outperformed overlapping windows, with the best accuracy of 56.11% achieved using PCA features with a 20-second window and a 0.1 learning rate. These findings indicate that eliminating temporal redundancy helps the model learn more robust and generalizable representations of cognitive load across individuals. Although adaptive learning rate strategy improved training stability, it did not surpass the performance of optimally selected fixed learning rates. The study highlights the critical role of segmentation strategy and learning rate selection in improving model generalization and identifies methodological considerations essential for developing reliable, real-time, and SI cognitive load classification systems using fNIRS.Keywords
This publication has 37 references indexed in Scilit:
- Cognitive load assessment through EEG: A dataset from arithmetic and Stroop tasksData in Brief, 2025
- Do automation and digitalization distract drivers? A systematic reviewAccident Analysis & Prevention, 2024
- Investigating construct validity of cognitive load measurement using single-item subjective rating scalesInstructional Science, 2024
- Unanticipated evolution of cardio-respiratory interactions with cognitive load during a Go-NoGo shooting task in virtual realityComputers in Biology and Medicine, 2024
- Exploring cognitive load through neuropsychological features: an analysis using fNIRS-eye trackingMedical & Biological Engineering & Computing, 2024
- Comparison of concurrent cognitive load measures during n-back tasksApplied Ergonomics, 2024
- Gaming behavior and brain activation using functional near‐infrared spectroscopy, Iowa gambling task, and machine learning techniquesBrain and Behavior, 2022
- Towards measuring cognitive load through multimodal physiological dataCognition, Technology & Work, 2020
- Computer Vision for Human–Machine Interaction2018
- Deep learning with convolutional neural networks for EEG decoding and visualizationHuman Brain Mapping, 2017