Wireless IoT networks are becoming increasingly important as they support a diverse range of applications. However, due to their resource constraints and limited security capabilities, these networks are particularly susceptible to attacks. Among these, jamming attacks pose a significant threat by degrading packet delivery, disrupting communications, and depleting the limited networking and energy resources of IoT systems. Hence, effective jamming detection is essential for safeguarding wireless IoT networks. However, dynamic operating environments and diverse application scenarios make it challenging to design a robust jamming detection system. To address the challenge, this paper introduces AJDet, a lightweight, online, Adaptive Jamming attack Detection system. AJDet leverages an online adaptation approach that enables the detection of both proactive and reactive jamming attacks without relying on a deployment-specific configuration. Our experimental results demonstrate that AJDet achieves over 96.5% detection accuracy without the need for manual intervention after deployment. Moreover, the system exhibits an efficient memory footprint, making it well-suited for resource-constrained devices.

In the ever-evolving adversarial machine learning landscape, developing effective defenses against patch attacks has become a critical challenge, necessitating reliable solutions to safeguard real-world AI systems. Although diffusion models have shown remarkable capacity in image synthesis and have been recently utilized to counter lp-norm bounded attacks, their potential in mitigating localized patch attacks remains largely underexplored. In this work, we propose DiffPAD, a novel framework that harnesses the power of diffusion models for adversarial patch decontamination. DiffPAD first performs super-resolution restoration on downsampled input images, then adopts binarization, dynamic thresholding scheme and sliding window for effective localization of adversarial patches. Such a design is inspired by the theoretically derived correlation between patch size and diffusion restoration error that is generalized across diverse patch attack scenarios. Finally, DiffPAD applies inpainting techniques to the original input images with the estimated patch region being masked. By integrating closed-form solutions for super-resolution restoration and image inpainting into the conditional reverse sampling process of a pre-trained diffusion model, DiffPAD obviates the need for text guidance or fine-tuning. Through comprehensive experiments, we demonstrate that DiffPAD not only achieves state-of-the-art adversarial robustness against patch attacks but also excels in recovering naturalistic images without patch remnants. The source code is available at https://github.com/JasonFu1998/DiffPAD. 

Introduction: Individuals recovering from COVID-19 often experience a range of post-recovery symptoms. However, the literature on post-COVID-19 symptoms reveals conflicting results, necessitating a heightened focus on longitudinal studies to comprehend the trajectory of impairments over time. Our study aimed to investigate changes in long-term impairments among individuals infected with COVID-19 and explore potential predictors influencing these changes. Methods: We conducted a web-survey targeting individuals that had been infected with COVID-19 at four time-points: T0 (baseline), T1 (three months), T2 (six months), and T3 (twelve months). The survey included contextual factors, factors related to body functions and structures, and post-COVID impairments. The longitudinal sample included 213 individuals (with a mean age of 48.92 years). Linear mixed models were employed to analyze changes in post-COVID impairments over time and identify impacting factors. Results: Findings revealed a general decline in post-COVID impairments over time, with each symptom exhibiting a dynamic pattern of fluctuations. Factors such as initial infection severity, education level, and work status were significantly associated with the levels of impairments. Discussion: The study emphasizes that post-COVID impairments are not static but exhibit variations over time. Personalized care, especially for vulnerable populations, is crucial. The results underscore the need for long-term monitoring and multidisciplinary treatment approaches. Targeted support and interventions are highlighted for individuals with severe initial infections and those in socioeconomically disadvantaged groups. 

The recent breakthroughs in machine learning (ML) and deep learning (DL) have catalyzed the design and development of various intelligent systems over wide application domains. While most existing machine learning models require large memory and computing power, efforts have been made to deploy some models on resource-constrained devices as well. A majority of the early application systems focused on exploiting the inference capabilities of ML and DL models, where data captured from different mobile and embedded sensing components are processed through these models for application goals such as classification and segmentation. More recently, the concept of exploiting the mobile and embedded computing resources for ML/DL model training has gained attention, as such capabilities allow the training of models via local data without the need to share data over wireless links, thus enabling privacy-preserving computation by design, model personalization and environment adaptation, and deployment of accurate models in remote and hardly accessible locations without stable internet connectivity. This work summarizes and analyzes state-of-the-art systems research that allows such on-device model training capabilities and provides a survey of on-device training from a systems perspective.

Background: The COVID-19 pandemic has triggered a global mental health crisis. Yet, we know little about the lasting effects of COVID-19 infection on mental health. This prospective longitudinal study aimed to investigate the trajectories of mental health changes in individuals infected with COVID-19 and to identify potential predictors that may influence these changes. Methods: A web-survey that targeted individuals that had been infected with COVID-19 was used at three time-points: T0 (baseline), T1 (six months), and T2 (twelve months). The survey included demographics, questions related to COVID-19 status, previous psychiatric diagnosis, post-COVID impairments, fatigue, and standardized measures of depression, anxiety, insomnia. Linear mixed models were used to examine changes in depression, anxiety, and insomnia over time and identify factors that impacted trajectories of mental health outcomes. Results: A total of 236 individuals completed assessments and was included in the longitudinal sample. The participants’ age ranged between 19 and 81 years old (M = 48.71, SD = 10.74). The results revealed notable changes in mental health outcomes over time. The trajectory of depression showed significant improvement over time while the trends in anxiety and insomnia did not exhibit significant changes over time. Younger participants and individuals who experienced severe COVID-19 infection in the acute phase were identified as high-risk groups with worst mental ill-health. The main predictors of the changes in the mental health outcomes were fatigue and post-COVID impairments. Conclusions: The findings of our study suggest that mental health outcomes following COVID-19 infection exhibit a dynamic pattern over time. The study provides valuable insights into the mental health trajectory following COVID-19 infection, emphasizing the need for ongoing assessment, support, and interventions tailored to the evolving mental health needs of this population. 

People living with type 1 diabetes often use several apps and devices that help them collect and analyse data for a better monitoring and management of their disease. When such health related data is analysed in the cloud, one must always carefully consider privacy protection and adhere to laws regulating the use of personal data. In this paper we present our experience at the pilot Vinter competition 2021-22 organised by Vinnova. The competition focused on digital services that handle sensitive diabetes related data. The architecture that we proposed for the competition is discussed in the context of a hypothetical cloud-based service that calculates diabetes self-care metrics under strong privacy preservation. It is based on Fully Homomorphic Encryption (FHE)-a technology that makes computation on encrypted data possible. Our solution promotes safe key management and data life-cycle control. Our benchmarking experiment demonstrates execution times that scale well for the implementation of personalised health services. We argue that this technology has great potentials for AI-based health applications and opens up new markets for third-party providers of such services, and will ultimately promote patient health and a trustworthy digital society.

Topic Detection (TD) refers to automatic techniques for locating topically related material in web documents. Nowadays, massive amounts of documents are generated by users of Online Social Networks (OSNs), in form of very short text, tweets and snippets of news. While topic detection, in its traditional form, is applied to a few documents containing a lot of information, the problem has now changed to dealing with massive number of documents with very little information. The traditional solutions, thus, fall short either in scalability (due to huge number of input items) or sparsity (due to insufficient information per input item). In this paper we address the scalability problem by introducing an efficient and scalable graph based algorithm for TD on short texts, leveraging dimensionality reduction and clustering techniques. We first, compress the input set of documents into a dense graph, such that frequent cooccurrence patterns in the documents create multiple dense topological areas in the graph. Then, we partition the graph into multiple dense sub-graphs, each representing a topic. We compare the accuracy and scalability of our solution with two state-of-the-art solutions (including the standard LDA, and BiTerm). The results on two widely used benchmark datasets show that our algorithm not only maintains a similar or better accuracy, but also performs by an order of magnitude faster than the state-of-the-art approaches.