This paper introduces a fusion convolutional architecture for efficient learning of spatio-temporal features in video action recognition. Unlike 2D convolutional neural networks (CNNs), 3D CNNs can be applied directly on consecutive frames to extract spatio-temporal features. The aim of this work is to fuse the convolution layers from 2D and 3D CNNs to allow temporal encoding with fewer parameters ...

<i>Background and objective</i>: The number of publications on the use of chatbots for health is recently increasing, however to our knowledge, there are no publications summarizing what is known about using chatbots for public health yet. The objective of this work is to provide an overview of the existing scientific literature on the use of chatbots for public health, for which purpose have chatbots ...

Accurate detection of accelerometer non-wear time is crucial for calculating physical activity summary statistics. In this study, we evaluated three epoch-based non-wear algorithms (Hecht, Troiano, and Choi) and one raw-based algorithm (Hees). In addition, we performed a sensitivity analysis to provide insight into the relationship between the algorithms’ hyperparameters and classification performance, ...

Pixel-wise image segmentation is a highly demanding task in medical-image analysis. In practice, it is difficult to find annotated medical images with corresponding segmentation masks. In this paper, we present Kvasir-SEG: an open-access dataset of gastrointestinal polyp images and corresponding segmentation masks, manually annotated by a medical doctor and then verified by an experienced ...

Highly automated vehicles will change our personal mobility in the future. To ensure the safety and the comfort of their passengers, the cars have to rely on as many information regarding their current surrounding traffic situation, as they can obtain. In addition to classical sensors like cameras or radar sensors, automated vehicles use data from a so called High Definition Street Map. Through such ...

Infection incidences in people with diabetes can create sever health complications mainly due to the effect of stress hormones, such as cortisol and adrenaline, which increases glucose production and insulin resistance in the body. The proposed electronic disease surveillance monitoring network (EDMON) relies on self-recorded data from people with Type 1 diabetes and dedicated algorithms to detect ...

Patients increasingly collect health-related data using mobile health apps and sensors. Studies have shown that this data can be beneficial for both clinicians and patients if used during medical consultations. However, such data is almost never used outside controlled situations or medical trials. This paper explains why the usage of self-collected health data is not widespread by identifying ...

Using smartphones for peer-to-peer communication over the Internet is difficult without the aid of centralized services. These centralized services, which usually reside in the cloud, are necessary for brokering communication between peers, and all communication must pass through them. A reason for this is that smartphones lack publicly reachable IP addresses. Also, because people carry their ...

Summary We report low baseline 25-hydroxyvitamin D (25OHD) levels being associated with unfavorable changes in muscle mass and physical performance over three years, but not with incident sarcopenia. Future prospective studies are needed to assess causality and to address the issue of competing risks such as mortality in older cohorts. Introduction Effects of low serum 25-hydroxyvitamin D (25OHD) ...

Quantification of training and match load is an important method to personalize the training stimulus’ prescription to players according to their match demands. The present study used time-motion analysis and triaxial-accelerometer to quantify and compare: a) The most demanding passages of play in training sessions and matches (5-min peaks); b) and the accumulated load of typical microcycles and ...