KMITL

Permanent URI for this communityhttps://dspace.kmitl.ac.th/handle/123456789/1

Browse

Search Results

Now showing 1 - 10 of 68

Some of the metrics are blocked by your
consent settings
Localized estimation of event-related neural source activity from simultaneous MEG-EEG with a recurrent neural network
(2024-12-01)
O'Reilly, Jamie A.
;
Zhu, Judy D.
;
Sowman, Paul F.
Estimating intracranial current sources underlying the electromagnetic signals observed from extracranial sensors is a perennial challenge in non-invasive neuroimaging. Established solutions to this inverse problem treat time samples independently without considering the temporal dynamics of event-related brain processes. This paper describes current source estimation from simultaneously recorded magneto- and electro-encephalography (MEEG) using a recurrent neural network (RNN) that learns sequential relationships from neural data. The RNN was trained in two phases: (1) pre-training and (2) transfer learning with L1 regularization applied to the source estimation layer. Performance of using scaled labels derived from MEEG, magnetoencephalography (MEG), or electroencephalography (EEG) were compared, as were results from volumetric source space with free dipole orientation and surface source space with fixed dipole orientation. Exact low-resolution electromagnetic tomography (eLORETA) and mixed-norm L1/L2 (MxNE) source estimation methods were also applied to these data for comparison with the RNN method. The RNN approach outperformed other methods in terms of output signal-to-noise ratio, correlation and mean-squared error metrics evaluated against reference event-related field (ERF) and event-related potential (ERP) waveforms. Using MEEG labels with fixed-orientation surface sources produced the most consistent estimates. To estimate sources of ERF and ERP waveforms, the RNN generates temporal dynamics within its internal computational units, driven by sequential structure in neural data used as training labels. It thus provides a data-driven model of computational transformations from psychophysiological events into corresponding event-related neural signals, which is unique among MEEG source reconstruction solutions.
Some of the metrics are blocked by your
consent settings
The Applications of Deep Learning in ECG Classification for Disease Diagnosis: A Systematic Review and Meta-Data Analysis
(2024-08-31)
Khalid, Mudassar
;
Pluempitiwiriyawej, Charnchai
;
Wangsiripitak, Somkiat
;
Murtaza, Ghulam
;
Abdulkadhem, Abdulkadhem A.
The supremacy of deep learning in artificial intelligence (AI) contexts, including image and speech recognition, computer vision, and medical imaging, among others, has established it as AI’s dominant approach. Several studies have been conducted on the use of deep learning in physiological signals, especially in ECG signals, in recent years, but there has been a lack of comprehensive review on the use of deep learning in ECG for biometric systems. This review is divided into two main sections: it provides a comprehensive bibliographic review of deep learning for ECG classification towards assisting in disease diagnosis in the first part while presenting an overview of the field, pioneers, and landmark studies. The second part offers comprehensive information on the subject, starting with the mathematical background of deep learning algorithms, the ECG signal processing, and the function of the heart. Using a PRISMA framework, 309 research papers were initially identified through specified keywords. After applying inclusion criteria, 90 articles were retained for detailed analysis, excluding 24 documents based on exclusion criteria EC1 and the remainder due to EC2. Key findings reveal that deep learning models achieve an average accuracy improvement of 10-15% over traditional methods, with convolutional neural networks (CNNs) and recurrent neural networks (RNNs) demonstrating superior performance in capturing complex ECG patterns. Through ECG databases, deep learning algorithms, assessment frameworks, metrics, and code availability, this review designs a systematic view from different perspectives to highlight the trends, challenges, and opportunities of deep learning for ECG arrhythmia classification. This paper’s goal is to contribute to the knowledge of both new and experienced researchers and practitioners in the field so that they can learn and understand the various processes involved in ECG signal processing using deep learning.
Some of the metrics are blocked by your
consent settings
ConvLSTM-based real-time power flow estimation of smart grid with high penetration of uncertain PV considering measurement noise
(2024-04-01)
Senesoulin, Fanta
;
Hongesombut, Komsan
;
Ngamroo, Issarachai
;
Dechanupaprittha, Sanchai
A modern smart grid tends to have increasingly various uncertain renewable generations. Due to different geographical areas and network topology constraints, operations of a smart grid become complicated and challenging. Moreover, using existing methods, power flow estimation in real-time could be time-consuming and computationally expensive. This paper proposes an efficient deep learning approach to estimate real-time power flow solutions of the smart grid with high penetration of uncertain PV generations using synchrophasor data considering measurement noise. The performance and effectiveness of the proposed convolutional long short-term memory (ConvLSTM) with time-series cross-validation technique are examined using synchrophasor data with Gaussian noise in the IEEE 39 bus test system. The proposed ConvLSTM approach shows better robust performance than weighted least square (WLS) state estimation and long short-term memory (LSTM) approaches. In addition, state measurements and confidence intervals are employed to confirm the accuracy of estimated real-time power flow results. The accurate real-time power flow estimation is crucial to determining dynamic available transfer capability (ATC) results and efficient operations of smart grids.
Some of the metrics are blocked by your
consent settings
Crack Localization Detection in Monolithic Zirconia Dental Crowns via 1D-Convolutional Neural Networks Algorithm-Based Acoustic Emission Analysis
(2024-01-01)
Tuntiwong, Kuson
;
Wangman, Rangsinee
;
Sritart, Hiranya
;
Kanchanatawewat, Kanchana
;
Tungjitkusolmun, Supan
The feasibility of utilizing the acoustic emission (AE) technique for the detection and classification of cracks within monolithic dental crown is assessed in this study, owing to its non-destructive nature which enables passive monitoring of structures. The AE signals captured are subjected to analysis to extract pertinent information regarding the source and location of the cracks. A novel approach is proposed, employing deep learning 1D convolutional neural networks (1D-CNNs) for the recognition and classification of the recorded cracked signals. The AE signals, obtained through a handmade AE data acquisition unit, are converted into .csv format and subjected to denoising using Bayesian methods to eliminate background noise. The signals are collected through the breakage of pencil lead (Vallen systeme) Hsu-Nielsen-Source 0.5 (ASTM E976) applied to each surface of the dental crown. Subsequently, the data signals are divided into training and testing groups following an 85 / 15 split. The performance of the deep learning 1D-CNNs is evaluated based on Precision, Recall and total accuracy metrics. The applicated of automated deep learning in this study demonstrated significantly high overall accuracy (98.67%). The integration of handmade data acquisition with 1D-CNN crack detection proves to be an effective method for early screening. The novel method harnesses acoustic emission signals in 1D-CNNs, thereby enhancing the accuracy of clinical dental restorative crack identification and determining the onset time.
Some of the metrics are blocked by your
consent settings
BAMS: Binary Sequence-Augmented Spectrogram with Self-Attention Deep Learning for Human Activity Recognition
(2024-01-01)
Sricom, Natchaya
;
Charakorn, Rujikorn
;
Manoonpong, Poramate
;
Limpiti, Tulaya
Human Activity Recognition (HAR) has rapidly gained interest over the years due to its wide range of applications in AI-based systems, particularly healthcare monitoring. HAR methods typically involve extracting relevant features from data provided by wearable sensors, smartphone sensors, cameras, or their combinations to classify different activities. Nevertheless, a major challenge lies in achieving high classification accuracy with limited data samples, particularly when distinguishing between activities with similar signal attributes. To address this challenge, we propose a novel HAR method called BinAry sequence-augmented spectrograM with Self-attention deep learning (BAMS). Our proposed method leverages only basic wearable sensor data. It utilizes short-time Fourier transform spectrograms to extract spatio-temporal sensor information. The spectrogram is integrated with a binary sequence that captures movement direction. We integrate a scaled dot-product self-attention mechanism into the model to prioritize data from wearable sensors, thereby enhancing the model's performance. The proposed method is evaluated on a public dataset using leave-one-subject-out cross-validation for efficacy and robustness. The method is found to achieve significant improvement over other state-of-the-art methods with the classification accuracy percentage and weighted F-1 scores of 88.06±5.11 and 87.36±5.96, respectively, for a twelve-activity classification.
Some of the metrics are blocked by your
consent settings
Spacing and Stressing Extracting of Sentences System
(2024-01-01)
Ngodngamjaras, Peerach
;
Teethawatthanakorn, Napatrapee
;
Anuntachai, Anuntapat
Word stressing and spacing to convey the meaning is an important part of speaking English. Audiences are more likely to engage with a speaker who uses proper word stress and spacing. Stress can be observed from louder pronunciation, and higher frequency. Space can be observed from breathing and the length of time before starting a new sentence. These things, if those who would like to practice English are inexperienced or do not listen to native speakers often, they may not be able to remember the points of stressing and spacing correctly. As a result, speech does not flow smoothly. Nowadays, there are no tutorials or visualizations that clearly point out where the spacing is. For this reason, the organizer decided to create a program that would identify word stress and spacing to display on a dashboard that would be developed into a website in the future.
Some of the metrics are blocked by your
consent settings
Enhanced Fire Detection Using Deep Learning and Heat Signatures
(2024-01-01)
Sinchai, Ananta
;
Pumanee, Ploychattra
;
Lomwong, Rattaphum
This study presents an innovative fire alarm system that integrates deep learning with thermal camera technology to tackle the urgent problem of slow fire detection, which often leads to considerable harm to individuals and extensive property damage. The system excels in identifying heat signatures prior to the full development of a fire, enabling timely notifications and thus mitigating potential damages and risks associated with fire incidents. Compared to traditional smoke detectors, this system operates at a considerably faster pace and offers a more flexible installation process by leveraging thermal cameras, which eliminates the need for ceiling-mounted detectors. Experimental evaluations demonstrate the efficacy of the proposed system, achieving up to 97% accuracy in fire detection. Simulations of various fire scenarios, ranging from initial heat detection to severe fire conditions, were used to train the system using a deep learning platform and a model of you only look once version 4 tiny (YOLOv4-Tiny). The results underscore the system's capability to detect early heat buildup swiftly, facilitating prompt alerts and enhancing overall fire safety.
Some of the metrics are blocked by your
consent settings
FILTER PRUNING BASED ON LOCAL GRADIENT ACTIVATION MAPPING IN CONVOLUTIONAL NEURAL NETWORKS
(2023-12-01)
Intraraprasit, Monthon
;
Chitsobhuk, Orachat
Convolutional Neural Network (CNN) is a well-known Deep learning model utilized extensively in the field of computer vision. The structure of convolutional neural networks is quite complicated and necessitates a substantial amount of computational time and storage resources. As a result, it is difficult to adopt a CNN model on a resource-constraint device. Model pruning can help to reduce computation time and storage re-quirements. In this research, we propose a filter pruning technique based on Localized Gradient Activation heatmaP (LGAP) for the purpose of pruning CNNs. Analyzing a filter based on statistical criterion of single neuron can lead to a loss in spatial relations within the filter activation itself, the relationship to target prediction, as well as the relationship among filters in that specific layer. To minimize the limitations, we evaluate the significance of a filter through the spatial information of local gradient activation related to the target prediction in terms of the layer-wise loss of the investigated filter. The effect of loss of an investigated filter demonstrates the significance or insignificance of the filter. Our pruning criteria ensure that these significant filters are preserved, while maintaining the model accuracy. The performance of our pruning method was validated using VGG-16 and ResNet-50. With pruning ratio of 50%, VGG-16 tends to decrease 1.66% of its accuracy, 3.6× of FLOP and 3.9× of storage reduction. For ResNet-50, with 50% pruning ratio, the results show that Top-1 and Top-5 of our pruning techniques outperform all the baseline techniques with a reduction of top-1 accuracy by 3.56%, top-5 accuracy by 1.89%, Floating Point Operation by 2.3×, and storage by 2.05×.
Some of the metrics are blocked by your
consent settings
Convolutional neural networks based focal loss for class imbalance problem: a case study of canine red blood cells morphology classification
(2023-11-01)
Pasupa, Kitsuchart
;
Vatathanavaro, Supawit
;
Tungjitnob, Suchat
Morphologies of red blood cells are normally interpreted by a pathologist. It is time-consuming and laborious. Furthermore, a misclassified red blood cell morphology will lead to false disease diagnosis and improper treatment. Thus, a decent pathologist must truly be an expert in classifying red blood cell morphology. In the past decade, many approaches have been proposed for classifying human red blood cell morphology. However, those approaches have not addressed the class imbalance problem in classification. A class imbalance problem—a problem where the numbers of samples in classes are very different—is one of the problems that can lead to a biased model towards the majority class. Due to the rarity of every type of abnormal blood cell morphology, the data from the collection process are usually imbalanced. In this study, we aimed to solve this problem specifically for classification of dog red blood cell morphology by using a Convolutional Neural Network (CNN)—a well-known deep learning technique—in conjunction with a focal loss function, adept at handling class imbalance problem. The proposed technique was conducted on a well-designed framework: two different CNNs were used to verify the effectiveness of the focal loss function and the optimal hyperparameters were determined by fivefold cross-validation. The experimental results show that both CNNs models augmented with the focal loss function achieved higher F<inf>1</inf> -scores, compared to the models augmented with a conventional cross-entropy loss function that does not address class imbalance problem. In other words, the focal loss function truly enabled the CNNs models to be less biased towards the majority class than the cross-entropy did in the classification task of imbalanced dog red blood cell data.
Some of the metrics are blocked by your
consent settings
Fingerprint-based Indoor Localization via Deep Learning
(2023-03-24)
Suroso, Dwi Joko
;
Cherntanomwong, Panarat
;
Sooraksa, Pitikhate
Deep learning (DL) application is proven helpful in a vast research field. One recent trend is to employ DL in radio frequency (RF)-based indoor localization. The fingerprint technique is the most used indoor localization technique known for its accuracy and performance. However, the fingerprint technique pays a high cost and effort in offline database construction, while its performance solely depends on the database density. Moreover, to apply deep learning, we also need a large dataset for it to learn efficiently. We propose to implement a DL-based fingerprint technique to tackle both problems of dataset scarcity and localization performance. We propose the DL's discriminative model, i.e., multilayer perceptron (MLP), for classification tasks. For the fingerprint database augmentation, we employed the generative model, i.e., Generative adversarial networks (GANs). We considered using a received signal strength indicator (RSSI) from a measurement campaign based on Wi-Fi devices for the database. The total area of interest is 25 m<sup>2</sup> inside the typical classroom environment, and we consider the 25 fingerprint locations as labels. We have a dataset of 1,250 rows x 8 columns (from 8 reference points). From the results, by using only 50% of actual data combined with the 125 synthetic data, we can improve the accuracy by more than 200% compared to only using 50% of actual data and show a 60% improvement in the loss. The combination of 100% actual data and 125 synthetic data gives the best accuracy and loss performance of 0.76 and 0.85, respectively. It gives an improvement of 144% in accuracy and 200% loss performance. By implementing deep learning for fingerprint techniques for data augmentation and classification, we can achieve good performance and reduce the workload of fingerprint database construction.

KMITL

Browse

Filters

Settings

Sort By

Results per page

Search Results