Rosula Reyes
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Rosula Reyes has a total of 61 co-authors. Here are the top co-authors by number of shared publications:
Samuel Matthew G. Dumlao
Department of Electronics, Computer and Communications Engineering, Ateneo de Manila University, Quezon City, Philippines
6 shared publications- Nicola Andriolli
University of Pisa
4 shared publications P. Castoldi
3 shared publications- Carlos Oppus
Electronics, Computer and Communications Engineering Department Ateneo de Manila University Quezon City, Philippines
2 shared publications - Charalambos Klitis
University of Glasgow
2 shared publications Derrick Sze
2 shared publications
Latest Publications
- Open Access
Non-invasive Diabetes Detection using Gabor Filter: A Comparative Analysis of Different Cameras
This paper compares and explores the performance of both mobile device camera and laptop camera as convenient tool for capturing images for non-invasive detection of Diabetes Mellitus (DM) using facial block texture features. Participants within age bracket 20 to 79 years old were chosen for the dataset. 12mp and 7mp mobile cameras, and a laptop camera were used to take the photo under normal lighting condition. Extracted facial blocks were classified using k-Nearest Neighbors (k-NN) and Support Vector Machine (SVM). 100 images were captured, preprocessed, filtered using Gabor, and iterated. Performance of the system was measured in terms of accuracy, specificity, and sensitivity. Best performance of 96.7% accuracy, 100% sensitivity, and 93% specificity were achieved from 12mp back camera using SVM with 100 images.
Open AccessNon-invasive Diabetes Detection using Gabor Filter: A Comparative Analysis of Different Cameras
This paper compares and explores the performance of both mobile device camera and laptop camera as convenient tool for capturing images for non-invasive detection of Diabetes Mellitus (DM) using facial block texture features. Participants within age bracket 20 to 79 years old were chosen for the dataset. 12mp and 7mp mobile cameras, and a laptop camera were used to take the photo under normal lighting condition. Extracted facial blocks were classified using k-Nearest Neighbors (k-NN) and Support Vector Machine (SVM). 100 images were captured, preprocessed, filtered using Gabor, and iterated. Performance of the system was measured in terms of accuracy, specificity, and sensitivity. Best performance of 96.7% accuracy, 100% sensitivity, and 93% specificity were achieved from 12mp back camera using SVM with 100 images.
- 27 November 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-8
A Non-intrusive Single Channel Abdominal Fetal Electrocardiogram Monitor Using Singular Value Decomposition
The maternal mortality rate (MMR) in the country is at 100 maternal deaths per 100,000 live births, which falls short from the Sustainable Development global goal of 70 maternal deaths per 100,000 live births. Current devices used to monitor fetal health are stationed at big hospitals, inaccessible to those living in remote areas. In this study, MAX-ECG-MONITOR, a device for collecting ECG signals, was used to develop a mobile application for fetal electrocardiogram (fECG) monitoring system. The signals gathered will be processed first through Singular Value Decomposition. The signal processing consists of three fundamental steps: estimation of maternal electrocardiogram (mECG), subtraction of maternal component from the waveforms, and extraction of fECG from the remaining signal. The signal processing was validated using a testing framework for the fetal ECG extraction algorithm. The program showed accuracy above 90 percent for signals with and without noise.
- 27 November 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-8
A Non-intrusive Single Channel Abdominal Fetal Electrocardiogram Monitor Using Singular Value Decomposition
The maternal mortality rate (MMR) in the country is at 100 maternal deaths per 100,000 live births, which falls short from the Sustainable Development global goal of 70 maternal deaths per 100,000 live births. Current devices used to monitor fetal health are stationed at big hospitals, inaccessible to those living in remote areas. In this study, MAX-ECG-MONITOR, a device for collecting ECG signals, was used to develop a mobile application for fetal electrocardiogram (fECG) monitoring system. The signals gathered will be processed first through Singular Value Decomposition. The signal processing consists of three fundamental steps: estimation of maternal electrocardiogram (mECG), subtraction of maternal component from the waveforms, and extraction of fECG from the remaining signal. The signal processing was validated using a testing framework for the fetal ECG extraction algorithm. The program showed accuracy above 90 percent for signals with and without noise.
- Open Access
- 1 February 2021
- Published by ASTES Journal in Advances in Science, Technology and Engineering Systems Journal
- Vol. 6 (1) , 1164-1170
Performance Evaluation of a Gamified Physical Rehabilitation Balance Platform through System Usability and Intrinsic Motivation Metrics
- 1 February 2021
- Published by ASTES Journal in Advances in Science, Technology and Engineering Systems Journal
- Vol. 6 (1) , 1164-1170
Open AccessPerformance Evaluation of a Gamified Physical Rehabilitation Balance Platform through System Usability and Intrinsic Motivation Metrics
- 7 May 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in Journal of Lightwave Technology
- Vol. 37 (16) , 3948-3955
Demonstration of a Multiplane OAM-Wavelength Packet Switch Controlled by a Two-Step Scheduler Implemented in FPGAs
In order to meet the demand for highly scalable and energy efficient data center switches, multiplane architectures, where multiple optical switching domains are leveraged, are an attractive solution. In this scenario, the orbital angular momentum (OAM) of light can be exploited as an additional domain along with wavelength, space, and time. In this work we report the demonstration of an OAM-wavelength photonic integrated packet switch controlled by an FPGA implementing a two-step scheduler suited for multiplane architectures. Packet-based bit error rate (BER) measurements show transmission at 20 Gb/s with BER $< 10^(-9)$ also for the most critical switch configurations, with penalty < 2 dB with respect to the back-to-back. The measurements demonstrate that the hardware-controlled switch can support up to 32 ports distributed in 4 cards with 8 ports each.- 7 May 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in Journal of Lightwave Technology
- Vol. 37 (16) , 3948-3955
Demonstration of a Multiplane OAM-Wavelength Packet Switch Controlled by a Two-Step Scheduler Implemented in FPGAs
In order to meet the demand for highly scalable and energy efficient data center switches, multiplane architectures, where multiple optical switching domains are leveraged, are an attractive solution. In this scenario, the orbital angular momentum (OAM) of light can be exploited as an additional domain along with wavelength, space, and time. In this work we report the demonstration of an OAM-wavelength photonic integrated packet switch controlled by an FPGA implementing a two-step scheduler suited for multiplane architectures. Packet-based bit error rate (BER) measurements show transmission at 20 Gb/s with BER $< 10^(-9)$ also for the most critical switch configurations, with penalty < 2 dB with respect to the back-to-back. The measurements demonstrate that the hardware-controlled switch can support up to 32 ports distributed in 4 cards with 8 ports each.- 1 January 2019
- Published by Optica Publishing Group
- p. W2A.35
First demonstration of an FPGA-controlled multiplane OAM-wavelength packet switch
We report the first demonstration of an OAM-wavelength photonic integrated packet switch, controlled by a FPGA-based scheduler suited for multiplane architectures. We show error-free transmission at 20 Gb/s in multiple OAM-wavelength configurations.
- 1 January 2019
- Published by Optica Publishing Group
- p. W2A.35
First demonstration of an FPGA-controlled multiplane OAM-wavelength packet switch
We report the first demonstration of an OAM-wavelength photonic integrated packet switch, controlled by a FPGA-based scheduler suited for multiplane architectures. We show error-free transmission at 20 Gb/s in multiple OAM-wavelength configurations.
- 1 October 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1381-1386
Aquaponics pH Level, Temperature, and Dissolved Oxygen Monitoring and Control System Using Raspberry Pi as Network Backbone
Monitoring environmental parameters in aquaponics is essential to sustain healthy living conditions for both fish and plants. In order to maximize growth for both, an optimal balance of environmental parameters present in the system is necessary. As a solution, an automated aquaponics system that reacts to parametric changes with an integrated web interface is developed. The Arduino microprocessor sends measured parameters to a central node which host a local server to store the data. Acquired data is processed and triggers corresponding actuators to maintain the optimal environmental parameters. All information are displayed in a mobile-friendly web application which enables access to the real-time and historical data, and manual controls of the actuators. Results shows that the calibrated sensors can gather sensor parameters with high accuracy through comparisons with commercial sensor readings. The control system is capable of running actuators automatically as a result of changes in sensor parameters or manually by the user. Each actuator was tested and successfully triggered for a set period of time for both manual and automated controls. Finally, the network backbone can transmit information and host a website that displays live sensor data, manually control the actuators, and save sensor reading logs. Sensor data are updated and are presented to the user every half seconds which ensures that the data presented are updated. A web application is also uploaded to Heroku which allows the user to view sensor logs and manually control the actuators anywhere. With the fully integrated system, the study was able to implement an automated aquaponics system that measures and controls pH, temperature, and dissolved Oxygen.
- 1 October 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1381-1386
Aquaponics pH Level, Temperature, and Dissolved Oxygen Monitoring and Control System Using Raspberry Pi as Network Backbone
Monitoring environmental parameters in aquaponics is essential to sustain healthy living conditions for both fish and plants. In order to maximize growth for both, an optimal balance of environmental parameters present in the system is necessary. As a solution, an automated aquaponics system that reacts to parametric changes with an integrated web interface is developed. The Arduino microprocessor sends measured parameters to a central node which host a local server to store the data. Acquired data is processed and triggers corresponding actuators to maintain the optimal environmental parameters. All information are displayed in a mobile-friendly web application which enables access to the real-time and historical data, and manual controls of the actuators. Results shows that the calibrated sensors can gather sensor parameters with high accuracy through comparisons with commercial sensor readings. The control system is capable of running actuators automatically as a result of changes in sensor parameters or manually by the user. Each actuator was tested and successfully triggered for a set period of time for both manual and automated controls. Finally, the network backbone can transmit information and host a website that displays live sensor data, manually control the actuators, and save sensor reading logs. Sensor data are updated and are presented to the user every half seconds which ensures that the data presented are updated. A web application is also uploaded to Heroku which allows the user to view sensor logs and manually control the actuators anywhere. With the fully integrated system, the study was able to implement an automated aquaponics system that measures and controls pH, temperature, and dissolved Oxygen.
- 1 October 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 2045-2049
Emotion Recognition via Facial Expression: Utilization of Numerous Feature Descriptors in Different Machine Learning Algorithms
Emotion Recognition has been a prominent study even before computers had the same computing power as of today. Human's emotions can be recognized through their body language, behavior and, most evidently, from the facial expression of the person. In facial image classification, each facial image can be represented through feature descriptors. Feature descriptors are simplified representations of the facial image that incorporates the essential key facial features. This study determines which feature descriptor will best fit a respective machine learning algorithm to classify facial expressions. Twelve possible combinations of Key Facial Detection, Saliency Mapping, Local Binary Pattern, and Histogram of Oriented Gradient are investigated together with six machine learning classification algorithms thus generating a total of seventy-two models. These will classify the following emotions: anger, disgust, fear, joy, neutral, sadness and surprise. A stratified ten-fold cross-validation is performed for verification on both the CK+ dataset and the locally gathered dataset for "in the wild" image testing. This study has determined that among the seventy-two models, the RBF SVM HOG+LBP model attained the highest average accuracy of 0.94 across the seven emotions with an F1 score of 0.93.
- 1 October 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 2045-2049
Emotion Recognition via Facial Expression: Utilization of Numerous Feature Descriptors in Different Machine Learning Algorithms
Emotion Recognition has been a prominent study even before computers had the same computing power as of today. Human's emotions can be recognized through their body language, behavior and, most evidently, from the facial expression of the person. In facial image classification, each facial image can be represented through feature descriptors. Feature descriptors are simplified representations of the facial image that incorporates the essential key facial features. This study determines which feature descriptor will best fit a respective machine learning algorithm to classify facial expressions. Twelve possible combinations of Key Facial Detection, Saliency Mapping, Local Binary Pattern, and Histogram of Oriented Gradient are investigated together with six machine learning classification algorithms thus generating a total of seventy-two models. These will classify the following emotions: anger, disgust, fear, joy, neutral, sadness and surprise. A stratified ten-fold cross-validation is performed for verification on both the CK+ dataset and the locally gathered dataset for "in the wild" image testing. This study has determined that among the seventy-two models, the RBF SVM HOG+LBP model attained the highest average accuracy of 0.94 across the seven emotions with an F1 score of 0.93.
- 1 April 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 287-290
Automated lung auscultation identification for mobile health systems using machine learning
An efficient classification system that aids in the computerized auscultation process was developed. A database of digital lung sounds was created from recorded lung sounds from anonymous patients using mobile application and digital stethoscopes. Efficiency of different classification algorithms to the dataset was tested, and their processing time was reduced up to 80.15% when applied with Principal Component Analysis (PCA). Among the six classification algorithms used, K-Nearest Neighbors (KNN) and Support Vector Machine (SVM) are more reliable to use in this dataset with a precision of 100% and 99.00%, respectively.
- 1 April 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 287-290
Automated lung auscultation identification for mobile health systems using machine learning
An efficient classification system that aids in the computerized auscultation process was developed. A database of digital lung sounds was created from recorded lung sounds from anonymous patients using mobile application and digital stethoscopes. Efficiency of different classification algorithms to the dataset was tested, and their processing time was reduced up to 80.15% when applied with Principal Component Analysis (PCA). Among the six classification algorithms used, K-Nearest Neighbors (KNN) and Support Vector Machine (SVM) are more reliable to use in this dataset with a precision of 100% and 99.00%, respectively.