Computer Science

Federated Learning (FL) is essential for building global models across distributed environments. However, it is significantly vulnerable to data and model poisoning attacks that can critically compromise the accuracy and reliability of the global model. These vulnerabilities become more pronounced in heterogeneous environments, where clients' data distributions vary broadly, creating a challenging setting for maintaining model integrity. Furthermore, malicious attacks can exploit this heterogeneity, manipulating the learning process to degrade the model or even induce it to learn incorrect patterns. In response to these challenges, we introduce RFCL, a novel Robust Federated aggregation method that leverages CLustering and cosine similarity to select similar cluster models, effectively defending against data and model poisoning attacks even amidst high data heterogeneity. Our experiments assess RFCL's performance against various attacker numbers and Non-IID degrees. The findings reveal that RFCL outperforms existing robust aggregation methods and demonstrates the capability to defend against multiple attack types.

It is becoming increasingly difficult to ignore the complexity of software products. Software metrics are proposed to help show indications for quality, size, complexity, etc. of software products. In this paper, software metrics related to complexity are developed and evaluated. A dataset of many open source projects is built to assess the value of the developed metrics. Comparisons and correlations are conducted among the different tested projects. A classification is proposed to classify software code into different levels of complexity. The results showed that measuring the complexity of software products based on decision coverage gives a significant indicator of degree of complexity of those software products. However, such indicator is not exclusive as there are many other complexity indicators that can be measured in software products. In addition, we conducted a comparison among several available metric tools that can collect software complexity metrics. Results among those different tools were not consistent. Such comparison shows the need to have a unified standard for measuring and collecting complexity attributes.

Medical imaging refers to several different technologies that are used to view the human body to diagnose, monitor, or treat medical conditions. It requires significant expertise to efficiently and correctly interpret the images generated by each of these technologies, which among others include radiography, ultrasound, and magnetic resonance imaging. Deep learning and machine learning techniques provide different solutions for medical image interpretation including those associated with detection and diagnosis. Despite the huge success of deep learning algorithms in image analysis, training algorithms to reach humanlevel performance in these tasks depends on the availability of large amounts of high-quality training data, including high-quality annotations to serve as ground-truth. Different annotation tools have been developed to assist with the annotation process. In this survey, we present the currently available annotation tools for medical imaging, including descriptions of graphical user interfaces (GUI) and supporting instruments. The main contribution of this study is to provide an intensive review of the popular annotation tools and show their successful usage in annotating medical imaging dataset to guide researchers in this area.

Medical imaging refers to several different technologies that are used to view the human body to diagnose, monitor, or treat medical conditions. It requires significant expertise to efficiently and correctly interpret the images generated by each of these technologies, which among others include radiography, ultrasound, and magnetic resonance imaging. Deep learning and machine learning techniques provide different solutions for medical image interpretation including those associated with detection and diagnosis. Despite the huge success of deep learning algorithms in image analysis, training algorithms to reach humanlevel performance in these tasks depends on the availability of large amounts of high-quality training data, including high-quality annotations to serve as ground-truth. Different annotation tools have been developed to assist with the annotation process. In this survey, we present the currently available annotation tools for medical imaging, including descriptions of graphical user interfaces (GUI) and supporting instruments. The main contribution of this study is to provide an intensive review of the popular annotation tools and show their successful usage in annotating medical imaging dataset to guide researchers in this area.

This study presents an assessment of walking stability in elderly people, focusing on local dynamic stability of walking. Its main objectives were to propose a technique to quantify local dynamic stability using nonlinear time-series analyses and a portable instrument, and to investigate their reliability in revealing the efficacy of an exercise training intervention for elderly people for improvement of walking stability. The method measured three-dimensional acceleration of the upper body, and computation of Lyapunov exponents, thereby directly quantifying the local stability of the dynamic system. Straight level walking of young and elderly subjects was investigated in the experimental study. We compared Lyapunov exponents of young and the elderly subjects, and of groups before and after the exercise intervention. Experimental results demonstrated that the exercise intervention improved local dynamic stability of walking. The proposed method was useful in revealing effects and efficacies of the exercise intervention for elderly people.

Researchers in computer science and computer engineering devote a significant part of their efforts on communication and interaction between man and machine. Indeed, with the advent of multimedia and multimodal processing in real time, the computer is no longer considered only as a computational tool, but as a machine for processing, communication, collection and control. Many machines assist and support many activities in daily life. The main objective of this paper is to propose a new methodological solution by modeling an architecture that facilitates the work of multimodal system especially for a fission module. To realize such systems, we rely on ontology to integrate data semantically. Ontologies provide a structured vocabulary usedas support for data representation. This paper provides a better understanding of the fission system and multimodal interaction. We present our architecture and the description of the detection of optimal modalities. This is done by using an ontological model that contains different applicable scenarios and describes the environment where a multimodal system exists. Fusion Fission Multimodal System Environment Environment Command Modality In1 Modality In2 Modality Inn Modality Ou1 Modality Ou2 Modality Oun Figure 1. Multimodal system. Copyright © 2013 SciRes. JSEA Modeling Rules Fission and Modality Selection Using Ontology 355

In pervasive computing, a set of smart devices communicate and collaborate together in order to provide adapted services to the user and applications, and help the former in his everyday life tasks. To adapt provided services, devices must be context-aware, which requires a good understanding and use of context. This term is still general, vague and doesn't help developer of context-aware applications in spite of the large number of proposed definitions. In this paper, we provide a service based definition of context that is abstract enough to be used in pervasive computing and limit the set of information required to make service adaptation. We use some scenarios to make clear our definition. We present also a new definition of context-awareness and a categorization of context based on the concept of service, which seems, to us, a fruitful approach in pervasive computing. Finally, we discuss our approach and contribution.

Using multimodality in a computing system is advantageous in the sense that it makes computing more accessible to a wide range of users including those with impairments. Our work is aimed at making informatics accessible to the visually impaired, specially the access to mathematical expressions. The selection of modalities, media, and presentation formats depends on the interaction context (i.e. combined contexts of the user, his environment and his computing system) and nature of the expression. The adaptation of a computing system to the needs of a mobile user is essential in order that the user could continue working on his task at anytime anywhere, thereby increasing his productivity. The system design is intended to be pervasive, fault-tolerant and having a great selfadaptation capability under varying conditions (e.g. missing or defective components, change in computing resources). This paper is aimed at presenting the modeling and the simulation of our system through case studies and formal specification using Petri nets. Also we propose ontology for context modeling. A simulation of the system's behavior, along with some test case scenarios, is also presented in this paper.

Ubiquitous (or Pervasive) Computing is a new domain in Computer Science resulting from the emergence and evolution of both distributed systems and mobile computing. Technology is moving beyond the personal computer towards a growing trend of embedded microprocessors in everyday objects and is demanding an unobtrusive connectivity between them in order to serve users at anytime and anywhere. The main objective of a ubiquitous computing system is to provide adaptive services proactively, without explicit user intervention and according to the user's current context. Despite interesting previous research works, there is still a lack of software tools and related research in terms of comprehensive context modeling, architecture of context-aware ubiquitous systems, and dynamic adaptation approaches in ubiquitous service computing environments. This chapter proposes a conceptual architecture to provide dynamic adaptability in ubiquitous services based on context-awareness and user preferences. As part of this proposal, the authors detail an ontology-based context modeling approach, a multi-agent architecture to support the development of ubiquitous computing applications, and a case-based reasoning method for service adaptation.