This paper introduces a motion planning approach for navigating in a dynamic environment. The path is represented using a Non-Uniform Rational B-Spline (NURBS) to ensure smoothness, curvature continuity, and proper orientation by adjusting its parameters. A Differential Evolution algorithm optimizes the curve parameters and traversal speed at each replanning interval, taking into account speed limits, maximum curvature, and obstacles in the environment. A constraintbased on Velocity Obstacle (VO) ensures collision-free motion, considering bounds provided by lower-level controllers. The feasibility of the approach is validated through simulations and real-world experiments with the Crazyflie 2.1 micro quadcopter.

Forest fires, often caused by extreme weather conditions or electrical failures, pose serious threats to property and human life. Rapid and accurate fire detection is essential for effective disaster response. In this study, we propose a real-time fire segmentation method for drone-captured video streams, balancing accuracy and computational efficiency. Our approach integrates YOLOv4 for fire detection and an Attention U-Net for precise segmentation. YOLOv4 first identifies fire regions using bounding boxes, which are then padded and refined through the Attention U-Net to achieve pixel-level segmentation. Various hyperparameters were evaluated for optimizing performance, including learning rate, batch size, and number of training epochs, with the objective of enhancing training diversity and improving generalization. Our results demonstrate the importance of hyperparameter tuning and model architecture in real-time fire detection. The proposed model effectively identifies large forest fires, improving early detection and response capabilities. This system has the potential to mitigate wildfire damage and save lives by enabling faster and more precise results.

Based on MRI images and a U-Net architecture, this study presents an optimized Convolutional Neural Network for segmenting Multiple Sclerosis (MS) lesions. The approach entails a two-step procedure designed to maximize accuracy in lesion detection. First, it identifies the most informative modality from the MRI images at hand. Then, it systematically assesses the critical hyper-parameters of the U-Net architecture to improve the learning process. By selecting the optimal modality and fine-tuning hyper-parameters, significant improvements in MS lesion detection and monitoring accuracy are achieved. This advancement facilitates a deeper understanding of the disease’s progression and its sub-types. The proposed approach holds promise for improving clinical decision-making and patient management in MS diagnosis and treatment.

Marfan syndrome is diagnosed by aortic enlargement, which can lead to complications such as aortic dilation, aortic dissection, and even rupture of the aortic wall in extreme cases. As part of this work, we developed a CNN system whose function is to interpret chest X-ray images and determine whether they show an enlarged aorta. Two architectures were studied for best hyper-parameters that affect the performance of DenseNet and Inception-v3. In most tests, the system obtained high accuracy rates, low loss rates, and no overfitting, making it a valuable tool for medical practitioners to help provide an initial diagnosis to potential patients.

The open-pit mining operational planning problem (OPMOP) is approached by applying Multi-Agent Systems (MAS) to optimize task distribution and find the most efficient paths simultaneously. Several agents are being investigated, including several trucks with different capabilities and operating costs, as well as shovels with differing production rates and compatibility. As open pit mines have a wide range of materials, shovels are assigned to them based on their characteristics. In addition, autonomous agents must be evaluated and optimized to use minimal resources while adhering to constraints. Simulations were done to assess how the system compares to human experience by using the IoT and online data to create a successful final schedule and plan.

This paper presents an approach for monitoring exercises of hand rehabilitation for past stroke patients. The developed solution uses a Leap Motion controller as hand-tracking device and embeds a supervised Machine Learning methodology. Support Vector Machine (SVM) is used in order to assess the correctness of a set of simple rehabilitation exercises performed with a single hand. The basic SVM model was extended with particular interest for defining feature vectors in a continues environment. The proposed method incorporated leap motion data, normalization of angles and gesture recognition. A software system was developed to provide patients with a set of exercise corrections and guidance for rehabilitation.

This paper describes the development of an SLR system for Israeli Sign Language designed to enable people with hearing impairments to communicate easily and independently using sign language. The research focuses on fingerspelling and the problem of acquiring and extracting data from signers and translating it into written form. SLR entails two main challenges. The first involves acquiring isolated instances of letters and translating each letter individually. This was accomplished by training the Support Vector Machines classifiers using vectors consisting of the Euclidean distances between the fingertips and the center of the palm. The second and more complex challenge involves translating a signed word and converting it into a stream of letters. This was solved in two sub-stages: extracting a single frame from the Leap controller and identifying the sign formed in that frame; and analyzing a continuous stream of frames and determining when the signs for a letter are formed. The recognition process in the system developed, eventually combined the letters to form words, moreover, provides a low-cost solution and is, therefore, suitable for everyone.