In this paper, we explore two modeling approaches to understanding the dynamics of infectious diseases in the population: equation-based modeling (EBM) and agent-based modeling (ABM). To achieve this, a comparative study of these approaches was conducted and we highlighted their advantages and disadvantages. Two case studies on the spread of the COVID-19 pandemic were carried out using both approaches. The results obtained show that differential equation-based models are faster but still simplistic, while agent-based models require more machine capabilities but are more realistic and very close to biology. Based on these outputs, it seems that the coupling of both approaches could be an interesting compromise.

Rapid and precise identification of Coronavirus Disease 2019 (COVID-19) is pivotal for effective patient care, comprehending the pandemic’s trajectory, and enhancing long-term patient survival rates. Despite numerous recent endeavors in medical imaging, many convolutional neural network-based models grapple with the expressiveness problem and overfitting, and the training process of these models is always resource-intensive. This paper presents an innovative approach employing Xception, augmented with cutting-edge transfer learning techniques to forecast COVID-19 from X-ray thorax images. Our experimental findings demonstrate that the proposed model surpasses the predictive accuracy of established models in the domain, including Xception, VGG-16, and ResNet. This research marks a significant stride toward enhancing COVID-19 detection through a sophisticated and high-performing imaging model.

Evacuating occupants from a burning building presents significant challenges owing to the complex interplay between environmental factors and evacuee behavior. Numerous models have been proposed in recent years for building evacuation during fire emergencies. However, many of these models fail to adequately account for variations in individuals’ environmental knowledge and health status, both of which can critically affect the evacuation process. To address this gap, this study introduces an agent-based model for building evacuation that differentiates between individuals familiar with the building layout and those who are not, and considers the health status of occupants. The resulting model assigned specific evacuation strategies tailored to each group within the building. Simulations conducted in a hospital setting demonstrate that incorporating factors such as familiarity with a building’s configuration and the health status of evacuees results in more realistic and effective evacuation outcomes. To evaluate the robustness of the proposed model and highlight the significance of our contribution, we conducted multiple simulations in various high-risk settings and presented a benchmarking study comparing our model with several existing models from the literature. The proposed model is sufficiently generalized to be applicable to various types of hospital buildings with minimal modifications.

Background: Given the vast amount of data generated daily on the Internet, numerous cryptosystems have been developed to ensure data confidentiality using symmetric, asymmetric, or hybrid encryption techniques. However, many of these systems suffer from limitations such as slow execution times and large key sizes. Objective: This paper presents a novel cryptosystem for text data encryption that integrates elliptic curve cryptography with chaotic functions to enhance data confidentiality and security. Methods: The proposed system utilizes elliptic curve points and pseudo-random numbers generated from a hybrid chaotic map (a fusion of the logistic map, sine map, and piecewise linear chaotic map) to derive short key sizes. Various analyses, including frequency histograms, correlation coefficients, and key sensitivity tests, are conducted to demonstrate the system’s robustness and reliability. Results: The analyses confirm that the fusion of chaotic functions makes the system highly sensitive to initial conditions, thus providing strong protection against unauthorized access. Conclusions: By integrating elliptic curve cryptography with chaotic functions, the proposed cryptosystem effectively addresses the limitations of existing systems, offering enhanced data confidentiality and security through the use of short key sizes and strong sensitivity to initial conditions.

Background: The regular use of connected devices on the internet by users of social media generates a large quantity of multimedia data every second, and the data concerned may be of a variable nature, such as text data, color images, video, and audio. We can confirm that online security is becoming an area that is increasingly being explored by researchers around the world to guarantee and ensure the authenticity and confidentiality of exchanged data. Objective: The goal of this article is to introduce a new dynamic cryptosystem for encrypting color images, using elliptic curves and chaotic functions. Methods: This new system serves to generate the coefficients of the curve, create the encryption keys, and create the prime number of the curve. The new system utilizes the length of the encryption keys, which encompasses all points on the curve and relates to the discrete logarithm problem of elliptic curves, thereby justifying the difficulty in uncovering hidden information. Results: The results were obtained with the assistance of several analyses, including histogram analysis, correlation coefficient analysis, differential attack analysis, and key space analysis. The results obtained confirm the robustness and reliability of the proposed cryptographic system. Conclusions: By combining the two systems (elliptic curves and chaotic functions), we obtain a new hybrid system capable of operating with reduced encryption key lengths and random values, which makes our new system sensitive to initial conditions.

Atmospheric fog poses a critical challenge for computer vision systems in autonomous driving, surveillance, and robotics, where reliable object classification is essential. Under severe fog, classification accuracy can degrade by over 50%, and most existing approaches rely on separate defogging steps, which limit their applicability in real-time settings. This study introduces the Enhanced Density-Aware Cross-Scale Transformer (EDCST), a novel architecture designed for direct object classification under foggy conditions without requiring prior defogging. To support model training and evaluation, we developed a physics-based simulation framework generating four fog types (uniform, gradient, patchy, and adaptive) across nine intensity levels. EDCST leverages 384 dimensional embeddings, eight transformer layers, and twelve attention heads, trained using curriculum learning and OneCycleLR scheduling. On CODaN-Fog (15,500 images at 224×224 resolution), EDCST achieves 84.4% accuracy on clean images and retains 74.2% accuracy under severe fog (80% intensity), outperforming baseline transformers by 15.8%. Class-wise sensitivity analysis reveals that larger objects, such as vehicles and animals, maintain over 75% classification performance, while smaller objects are more affected. Patchy fog causes the greatest accuracy drop (19.1%), followed by adaptive (8.9%) and uniform fog (6.8%). The model converges in 100 epochs within 513 minutes. This work introduces a real-time-capable classification framework that eliminates defogging requirements and maintains strong performance under diverse fog conditions, making it highly suitable for safety-critical vision applications.