Search

Mineral grain recognition is an extremely important task in many fields, especially in mineral exploration, when trying to identify locations where precious minerals can possibly be found. The usual manual method would be to collect samples; a specialized individual using expensive equipment would manually identify and then count the grain minerals in the sample. This is a tedious task that is time-consuming and expensive. It is also limited because small portions of areas can be surveyed; even then, it might require extremely long periods. In addition, this process is still prone to human errors. Developing an automatic system to identify, recognize, and count grain minerals in samples from images would allow for more precise results than the time required by humans. In addition, such systems can be fitted on robots that collect samples, take images of the samples, and then proceed with the automated recognition and counting algorithm without human intervention. Vast amounts of land can be surveyed in this way. This paper proposes a modified approach for microscopic grain mineral recognition and classification using hybrid features and ensemble algorithms from images. The enhanced approach also included a modified segmentation approach, which enhanced the results. For 10 classes of microscopic mineral grains, using the modified approach and the ensemble algorithm resulted in an average accuracy of 84.01%. For 8 classes, the average reported accuracy is 94.93% using the Boosting ensemble learning with the C4.5 classifier. The results obtained outperform similar methods reported in the extant literature. [ABSTRACT FROM AUTHOR]

Numerous studies have demonstrated the adverse impact of the COVID-19 global pandemic on the mental health of post-secondary students worldwide. However, there have been relatively few investigations conducted in a university setting that encompasses both students and employees. Furthermore, almost all previous studies relied on conventional statistical analysis. Our research specifically examined anxiety levels among the Quebec university community during the COVID-19 pandemic, with a focus on the Generalized Anxiety Disorder (GAD-7) score using both conventional data exploration and predictive machine learning techniques. We found that the CatBoost algorithm was the best predictor of the GAD-7 score, achieving a squared Pearson correlation coefficient of r 2 = 0.5656 . Additionally, we utilized SHapley Additive exPlanations (SHAP) to explore variable importance and interaction effects between the variables in the predictive model. Finally, we also examined anxiety levels from the perspective of a classification task. Unfortunately, the results obtained were not as satisfactory as those obtained by the regression approach. [ABSTRACT FROM AUTHOR]

Human activity recognition (HAR) is among the most popular research topics. Indeed, recognizing human activities can help provide appropriate assistance to older adults and address the challenges of an aging population. Hence, HAR solutions based on Ambient Intelligence (AmI) have been proposed to face the challenges of an aging population. In addition, we denoted an increasing interest in Ultra-WideBand (UWB) radars for HAR. In this work, we exploited three UWB radars to recognize 15 human activities performed by 9 participants in a prototype apartment. One of the main contributions of this paper is the improvement of classification results compared to our previous work, with an average accuracy of 26% higher for the top-1 classification. To do so, we emphasize the data cleaning stage. More precisely, since the amount of data is insignificant, data provided by UWB radars have been cleaned with a well-known band-pass Chebyshev type I filter of order 2. Applying that kind of filter to data provided by UWB radars is uncommon in the literature. In addition, two different deep learning architectures to classify the cleaned data have been exploited. The first is a relatively simple Convolutional Neural Network (CNN), and the second is Efficient-CapsNet. We obtained similar performances between these two architectures for the top-1 with an accuracy of approximately 64%. However, from the top-2 to top-5, the CNN outperformed Efficient-CapsNet. [ABSTRACT FROM AUTHOR]

In this paper, we propose new deep learning architectures to fuse data provided by multiple sensors. More specifically, we combine classical features extracted from a sensor and raw data of other sensors. In order to make this data fusion possible, we exploited convolution, dense, and concatenation layers. The Mobile HEALTH dataset has been used to support our study. The results show that the proposed architectures are suitable for future use in the Human Activity Recognition (HAR) domain since their performances are comparable or better than those presented in the recent literature and the reference architectures. Indeed, we reached approximately an accuracy of 0.965 and 0.995 for the leave-one-subject-out and train-test strategies, respectively. [ABSTRACT FROM AUTHOR]

Due to its sturdiness, low cost and ease of implementation, the induction machine is one of the most common electrical motors used in industry. However, this machine still concedes failures leading to unplanned shutdowns, sources of significant financial losses. For this reason, induction machine failure diagnosis has become an ordinary task, where some ameliorations have to be achieved in order to improve the efficiency of maintenance programs by minimizing the number and the time period of unexpected shutdowns. In this paper, we address the recognition of inter-turn short circuit (ITSC) in the stator windings and broken rotor bars (BRBs) in the three-phase squirrel cage induction machine by proposing a new approach. This approach relies on the three-phase current analysis method to extract features and a new hierarchical recognition algorithm based on an ensemble of three different classifiers. In addition, the present work studies the failures recognition of the induction machine operating with a torque and speed control. The proposed approach is a major advanced for the induction machine diagnosis research because it recognizes correctly (detects, localizes and estimates the degree of severity) ITSC and (detects and estimates) BRB faults with an accuracy of 93.36% and with a great robustness compared to a classical machine learning algorithm. [ABSTRACT FROM AUTHOR]

The purpose of the study was to compare the effects of sensory manipulations on postural control for subjects of different ages. A young group of subjects ( n = 17; 20.0 ± 1.3 years) and an old group of subjects ( n = 17; 74.7 ± 6.3 years) were compared in 14 postural conditions [2 reference conditions and 12 sensory manipulation conditions: eyes closed, cervical collar, tendon vibration, electromyostimulation, galvanic vestibular stimulation (2 designs), foam surface] on a force platform. Spatio-temporal parameters of the center of foot pressure displacement were analyzed. When vestibular or proprioceptive afferences were manipulated, the old group was more disturbed than the young group. In addition, when myo-articular proprioceptive afferences were the only non-manipulated information source, the old group was also more disturbed than the young group. Hence, the inability to correctly interpret proprioceptive information and/or the impairment of myo-articular information would appear to be the major factor causing postural control deterioration. Moreover, concerning the vestibular system, it may be that aging alters the central integration of vestibular afferences. These results suggest that aging differently affects the functional ability of the different neural loops in postural control. [ABSTRACT FROM AUTHOR]

The aim of this work was to compare the effects of fatigue of the quadriceps femoris after fatiguing voluntary contractions (VOL) and fatiguing neuromuscular electrical stimulation (ES) on bipedal postural control. Nineteen active male subjects (22.2 ± 1.7 years) completed these two fatiguing exercises. Isometric maximal voluntary contraction (MVC) and postural control were recorded using an ergometer and a force platform that registered the center of foot pressure (COP). We analyzed the COP surface, the mean COP velocity and the spectral power density given by the wavelet transform. Recordings were performed before (PRE condition) and after the completion of each fatiguing task (immediately POST condition, after a 5 min recovery POST 5 condition). In POST condition, the ES exercise affected MVC more than the VOL exercise. However, bipedal postural control was similarly deteriorated for both exercises. In POST 5 condition, for both fatiguing exercises, muscle strength and postural control did not recover their initial level. These results suggest that the postural control disturbance could not be distinguished for the two fatiguing exercises in the bipedal stance. In addition, the recovery speeds of postural control and muscle strength abilities did not differ for the ES exercise and the VOL exercise. [ABSTRACT FROM AUTHOR]