Mental Workload Classification and Tasks Detection in Multitasking: Deep Learning Insights from EEG Study

Brain Sciences, Journal Year: 2024, Volume and Issue: 14(2), P. 149 - 149

Published: Jan. 31, 2024

While the term task load (TL) refers to external demands, amount of work, or number tasks be performed, mental workload (MWL) individual’s effort, capacity, cognitive resources utilized while performing a task. MWL in multitasking scenarios is often closely linked with quantity person handling within given timeframe. In this study, we challenge hypothesis from perspective electroencephalography (EEG) using deep learning approach. We conducted an EEG experiment 50 participants NASA Multi-Attribute Task Battery II (MATB-II) under 4 different levels. designed convolutional neural network (CNN) help two distinct classification tasks. one setting, CNN was used classify segments based on their level. another same architecture trained again detect presence individual MATB-II subtasks. Results show that, model successfully learns whether particular subtask active segment (i.e., differentiate between subtasks-related patterns), it struggles highest levels distinguish MWL-related patterns). speculate that comes factors: first, way these differed only work timeframe; and second, participants’ effective adaptation increased as evidenced by low error rates. Consequently, indicates such conditions multitasking, may not reflect enough patterns higher load.

Language: Английский

---

A novel approach of brain-computer interfacing (BCI) and Grad-CAM based explainable artificial intelligence: Use case scenario for smart healthcare

Journal of Neuroscience Methods, Journal Year: 2024, Volume and Issue: 408, P. 110159 - 110159

Published: May 7, 2024

Language: Английский

---

Quantifying Soybean Defects: A Computational Approach to Seed Classification Using Deep Learning Techniques

Agronomy, Journal Year: 2024, Volume and Issue: 14(6), P. 1098 - 1098

Published: May 22, 2024

This paper presents a computational approach for quantifying soybean defects through seed classification using deep learning techniques. To differentiate between good and defective seeds quickly accurately, we introduce lightweight defect identification network (SSDINet). Initially, the labeled dataset is developed processed proposed contour detection (SCD) algorithm, which enhances quality of images performs segmentation, followed by SSDINet. The network, SSDINet, consists convolutional neural depthwise convolution blocks, squeeze-and-excitation making lightweight, faster, more accurate than other state-of-the-art approaches. Experimental results demonstrate that SSDINet achieved highest accuracy, 98.64%, with 1.15 M parameters in 4.70 ms, surpassing existing models. research contributes to advancing techniques agricultural applications offers insights into practical implementation systems control industry.

Language: Английский

---

Empowering drones in vehicular network through fog computing and blockchain technology

PLoS ONE, Journal Year: 2025, Volume and Issue: 20(1), P. e0314420 - e0314420

Published: Jan. 24, 2025

The performance of drones, especially for time-sensitive tasks, is critical in various applications. Fog nodes strategically placed near IoT devices serve as computational resources ensuring quick service responses deadline-driven tasks. However, the limited battery capacity drones poses a challenge, necessitating energy-efficient Internet Drones (IoD) systems. Despite increasing demand drone flying automation, there significant absence comprehensive network architecture tailored secure and efficient operations drones. This research paper addresses this gap by proposing safe, reliable, real-time architecture, emphasizing collaboration with fog computing. contribution includes systematic design integration blockchain technology data storage. computing was introduced Drone Blockchain Technology (FCDBT) model, where collaborate to process efficiently. proposed algorithm dynamically plans trajectories optimizes computation offloading. Results from simulations demonstrate effectiveness showcasing reduced average response latency improved throughput, particularly when accessing nodes. Furthermore, model evaluates consensus algorithms (PoW, PoS, DAG) recommends DAG superior handling data. Fog; Drones; Blockchain; PSO; IoT; Vehicular.

Language: Английский

---

Brain–computer interfaces and deep learning methods for cognitive impairments

Elsevier eBooks, Journal Year: 2025, Volume and Issue: unknown, P. 145 - 159

Published: Jan. 1, 2025

Language: Английский

---

Flexible wearable electronics for enhanced human-computer interaction and virtual reality applications

Nano Energy, Journal Year: 2025, Volume and Issue: 138, P. 110821 - 110821

Published: March 5, 2025

Language: Английский

---

Efficient Approach for Brain Tumor Detection and Classification Using Fuzzy Thresholding and Deep Learning Algorithms

IEEE Access, Journal Year: 2025, Volume and Issue: 13, P. 78808 - 78832

Published: Jan. 1, 2025

Language: Английский

---

AMFTCNet: A multi-level attention-based multi-scale fusion temporal convolutional network for decoding MI-EEG signals

Biomedical Signal Processing and Control, Journal Year: 2025, Volume and Issue: 108, P. 107916 - 107916

Published: May 1, 2025

---

Feature Fusion for Improved Classification: Combining Dempster-Shafer Theory and Multiple CNN Architectures

Lecture notes in computer science, Journal Year: 2024, Volume and Issue: unknown, P. 280 - 292

Published: Jan. 1, 2024

Language: Английский

---

Effective Detection of Epileptic Seizures through EEG Signals Using Deep Learning Approaches

Machine Learning and Knowledge Extraction, Journal Year: 2023, Volume and Issue: 5(4), P. 1937 - 1952

Published: Dec. 11, 2023

Epileptic seizures are a prevalent neurological condition that impacts considerable portion of the global population. Timely and precise identification can result in as many 70% individuals achieving freedom from seizures. To achieve this, there is pressing need for smart, automated systems to assist medical professionals identifying disorders correctly. Previous efforts have utilized raw electroencephalography (EEG) data machine learning techniques classify behaviors patients with epilepsy. However, these studies required expertise clinical domains like radiology procedures feature extraction. Traditional classification relied on manual engineering, limiting performance. Deep excels at directly sans human effort. For example, deep neural networks now show promise analyzing EEG detect seizures, eliminating intensive or engineering needs. Though still emerging, initial demonstrate practical applications across domains. In this work, we introduce novel residual model called ResNet-BiGRU-ECA, brain activity through accurately identify epileptic evaluate our proposed model’s efficacy, used publicly available benchmark dataset The results experiments demonstrated suggested surpassed both basic cutting-edge models, an outstanding accuracy rate 0.998 top F1-score 0.998.

Language: Английский

---