Lab

SDAS Research Group

About the lab

Smart Data Analysis Systems Group (SDAS Group) is supported by experts and young researchers from several academic institutions from different countries.

We develop software- and intelligent-systems-driven solutions for data-analysis-related problems of industry, medicine and education fields.

By vocation, all our R&D processes have a strong and genuine commitment to the synergistic, ethical and interactive integration of human thinking and artificial intelligence.

A wide range of data analysis related topics are covered by our research programs (https://sdas-group.com/researchtopics/).

At a technical consulting and deployment level, SDAS Group offers specialized services.

Featured projects (1)

Project
This researh program covers the following areas: - Pattern recognition and data mining. - Time series forecasting. - Dimensionality reduction and data visualization. - Clustering. - Learning from multiple experts. - Case-based reasoning. - Spectral, graph-based and kernel-based approaches for data analysis. - Information quality and data fusion. - Dynamic data analysis. Find more info at: https://sdas-group.com/machine-learning/

Featured research (57)

Automatic crop classification using new technologies is recognized as one of the most important assets in today’s smart farming improvement. Investments in technology and innovation are key issues for shaping agricultural productivity as well as the inclusiveness and sustainability of the global agricultural transformation. Digital image processing (DIP) has been widely adopted in this field, by merging Unmanned Aerial Vehicle (UAV) based remote sensing and deep learning (DL) as a powerful tool for crop classification. Despite the wide range of alternatives, the proper selection of a DL approach is still an open and challenging issue. In this work, we carry out an exhaustive performance evaluation of three remarkable and lightweight DL approaches, namely: Visual Geometry Group (VGG), Residual Neural Network (ResNet) and Inception V3, tested on high resolution agriculture crop images dataset. Experimental results show that InceptionV3 outperforms VGG and ResNet in terms of precision (0,92), accuracy (0,97), recall (0,91), AUC (0,98), PCR (0,97), and F1 (0,91).
Genetic algorithms (GAs) are a great tool for solving optimization problems. Their characteristics and different components based on the principles of biological evolution make these algorithms very robust and efficient in this type of problem. Many research works have presented dedicated solutions to schedule or resource optimization problems in different areas and project types; most of them have adopted GA implementation to find an individual that represents the best solution. Under this conception, in this work, we present a GA with a controlled mutation operator aiming at maintaining a trade-off between diversity and survival of the best individuals of each generation. This modification is supported by an improvement in terms of convergence time, efficiency of the results and the fulfillment of the constraints (of 29%, 14.98% and 23.33% respectively, compared with state-of-the-art GA with a single random mutation operator) to solve the problem of schedule optimization in the use of three laboratory rooms of the Mechatronics Engineering Career of the International University of Ecuador.
In fields such as hydrology, meteorology, and civil engineering, the study of extreme precipitation events is useful to prevent rainfall related disasters. A widely-used practice to address such a problem is by using statistical inferences about precipitation intensity, duration and frequency (IDF). Despite of its great usefulness, the selection of the adequate data and methodology to characterize precipitation’s IDF in the urban area of high-altitude Andean cities remains an open issue for practitioners and decision makers. In this sense, the present paper develops an approach to schematically build the IDF curves for a sub-basin of the study case Andean city, Quito–Ecuador. The here-used data holds information from 12 meteorological stations. Then, the IDF curves are obtained by using both a parametrization followed by a Gamma distribution and a 3-parameter cumulative distribution function, also called mnp. Finally, the curve-fitting process is estimated numerically by adjusting the Sherman equation. Results (average R2=0.9) demonstrated that the framework is well-suited for the high-altitude regime. As a noticeable outcome, a novel spatial interpolation-based analysis is introduced, which enabled the identification of extreme rainfall events according to its duration.
Real-time vital signs monitoring, particularly heart rate, is essential in today's medical practice and research. Heart rate detection allows the doctor to monitor the patient's health status to provide immediate action against possible cardiovascular diseases. We present a possible alternative to traditional heart rate signal monitoring systems, a cardiac pulse system using low-cost piezoelectric signal identification. This system could benefit health care and develop continuous pulse waveform monitoring systems. This paper introduces a heartbeat per minute (BPM) cardiac pulse detection system based on a low-cost piezoelectric ceramic sensor (PCS). The PCS is placed under the wrist and adjusted with a silicone wristband to measure the pressure exerted by the radial artery on the sensor and thus obtain the patient's BPM. We propose a signal conditioning stage to reduce the sensor's noise when acquiring the data and make it suitable for real-time BPM visualization. As a comparison, we performed a statistical test to compare the low-cost PCS with types of traditional sensors, along with the help of 21 volunteers. Experimental results show that the data collected by the PCS, when used for heart rate detection, is highly accurate and close to traditional sensor measurements. Therefore, we conclude that the system efficiently monitors the cardiac pulse signal in BPM. Keywords: Heart rate; Piezoelectric, BPM; Pulse Detection.
Face mask detection has become a great challenge in computer vision, demanding the coalition of technology with COVID-19 awareness. Researchers have proposed deep learning models to detect the use of face masks. However, the incorrect use of a face mask can be as harmful as not wearing any protection at all. In this paper, we propose a compound convolutional neural network (CNN) architecture based on two computer vision tasks: object localization to discover faces in images/videos, followed by an image classification CNN to categorize the faces and show if someone is using a face mask correctly, incorrectly, or not at all. The first CNN is built upon RetinaFace, a model to detect faces in images, whereas the second CNN uses a ResNet-18 architecture as a classification backbone. Our model enables an accurate identification of people who are not correctly following the COVID-19 healthcare recommendations on face mask use. To enable further global use of our technology, we have released both the dataset used to train the classification model and our proposed computer vision pipeline to the public, and optimized it for embedded systems deployment.

Lab head

Diego Peluffo
Department
  • Research Board
About Diego Peluffo
  • He received his degree in electronic engineering, the M.Eng.and PhD. degree in industrial automation from the Universidad Nacional de Colombia, Manizales - Colombia, in 2008, 2010 and 2013, respectively. He undertook his doctoral internship at KU Leuven - Belgium. Afterwards, he worked as a post-doc at Université Catholique de Louvain at Louvain la-Neuve, Belgium. He is the head of the SDAS Research Group.

Members (35)

Miguel A Becerra
  • Instituto Tecnologico Pascual Bravo
Paul Rosero
  • IT University of Copenhagen
Ana Cristina Umaquinga
  • Universidad Técnica del Norte
Leandro Lorente
  • SDAS Research Group
Dagoberto Mayorca
  • Universidad Mariana
Israel Herrera
  • Universidad Técnica del Norte

Alumni (2)

Andres Javier Anaya Isaza
  • Pontificia Universidad Javeriana
Diana M. Viveros Melo
  • Pontifícia Universidade Católica do Rio de Janeiro