In Federated Learning (FL), a strong global model is collaboratively learned by aggregating the clients' locally trained models. Although this allows no need to access clients' data directly, the global model's convergence often suffers from data heterogeneity. This paper suggests that forgetting could be the bottleneck of global convergence. We observe that fitting on biased local distribution shifts the feature on global distribution and results in forgetting of global knowledge. We consider this phenomenon as an analogy to Continual Learning, which also faces catastrophic forgetting when fitted on the new task distribution. Based on our findings, we hypothesize that tackling down the forgetting in local training relives the data heterogeneity problem. To this end, we propose a simple yet effective framework Federated Local Self-Distillation (FedLSD), which utilizes the global knowledge on locally available data. By following the global perspective on local data, FedLSD encourages the learned features to preserve global knowledge and have consistent views across local models, thus improving convergence without compromising data privacy. Under our framework, we further extend FedLSD to FedLS-NTD, which only considers the not-true class signals to compensate noisy prediction of the global model. We validate that both FedLSD and FedLS-NTD significantly improve the performance in standard FL benchmarks in various setups, especially in the extreme data heterogeneity cases.Preprint. Under review.
Individuals who are unable to walk independently spend most of the day in a wheelchair. This population is at high risk for developing pressure injuries caused by sitting. However, early diagnosis and prevention of these injuries still remain challenging. Herein, we introduce battery-free, wireless, multimodal sensors and a movable system for continuous measurement of pressure, temperature, and hydration at skin interfaces. The device design includes a crack-activated pressure sensor with nanoscale encapsulations for enhanced sensitivity, a temperature sensor for measuring skin temperature, and a galvanic skin response sensor for measuring skin hydration levels. The movable system enables power harvesting, and data communication to multiple wireless devices mounted at skin-cushion interfaces of wheelchair users over full body coverage. Experimental evaluations and numerical simulations of the devices, together with clinical trials for wheelchair patients, demonstrate the feasibility and stability of the sensor system for preventing pressure injuries caused by sitting.
Considerable efforts have been devoted to developing wound dressings with various functions, including rapid cell proliferation, protection against infection, and wound state monitoring to minimize severe pain and the risks of wound‐caused secondary infections. However, it remains challenging to diagnose wound conditions and achieve integration of the above functions without specialized equipment and expertise in wound care. This study describes an electrospun composite micro/nanofiber‐based bilayer‐dressing patch comprising a healing‐support layer (hyaluronic acid, gelatin, and dexpanthenol) and a protective/monitoring layer (curcumin and polycaprolactone). The improved cell regeneration function and biocompatibility of the healing‐support layer enable rapid healing, as evidenced by the expedited growth of fibroblasts. The superior antimicrobial properties (against Escherichia coli and Staphylococcus aureus) and visible color changes within the pH range of wound lesions (pH 6–9) of the protective/monitoring layer make the dressing suitable for advanced wound care. The wounds inflicted on BALB/c mice heal rapidly (12 days) without scars while the wound state can be diagnosed by the change in color of the dressing patch. The multifunctional wound dressing patch developed in this study is expected to promote wound healing and monitor wound state; thus, facilitating convenient wound management.
Recently, artificial intelligence has been successfully used in fields, such as computer vision, voice, and big data analysis. However, various problems, such as security, privacy, and ethics, also occur owing to the development of artificial intelligence. One such problem are deepfakes. Deepfake is a compound word for deep learning and fake. It refers to a fake video created using artificial intelligence technology or the production process itself. Deepfakes can be exploited for political abuse, pornography, and fake information. This paper proposes a method to determine integrity by analyzing the computer vision features of digital content. The proposed method extracts the rate of change in the computer vision features of adjacent frames and then checks whether the video is manipulated. The test demonstrated the highest detection rate of 97% compared to the existing method or machine learning method. It also maintained the highest detection rate of 96%, even for the test that manipulates the matrix of the image to avoid the convolutional neural network detection method.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.