No description available.

No description available.

No description available.

No description available.

Football fans do not have the time to watch every game in its entirety and need an effective solution that summarizes them the story of the game. Human-generated summaries are often too short, requiring time and resources to create. We utilize the advantages of Interactive TV to create an automatic football summarization service that is cohesive, provides context, covers the necessary plays, and is concise. First, we construct a degree of interest function that ranks each play based on detailed, play-by-play game events as well as viewing statistics collected from an interactive viewing environment. This allows us to select the plays that are important to the game as well as those that are interesting to the viewer. Second, we create a visual transition that shows the progress of the ball whenever plays are skipped, allowing the viewer to understand the context of each play within the summary. Third, we enable interactive controls that allow viewers to manipulate the summary and delve deeper into the actual game whenever they wish. We validate our solution through two user studies—one to ensure that our degree of interest function selects the plays that are most interesting to the viewer, and the other to show that our transitions and interactive controls provide a better understanding of the game. We conclude that our summary solution is effective at conveying the story of a football game.

The increasing use of health information technology (HIT) is due to a rising interest in improving the quality of health care. HIT has the potential to reduce cost and transform services. Proper clinical support systems will contribute to the meaningful use of HIT systems by providing a wide array of data to clinicians for the diagnosis and treatments. Clinical guidelines, created by a consensus of experts, can be put in place to assist physicians in making clinical decisions. Delphi methods are commonly used to create consensus from surveys completed by a team of experts. Image-based studies could create guidelines that standardize severity, deformity or other clinical classifications. As these studies were traditionally conducted using paper-based media, the cost and time requirement often make the process impractical. Using state of the art Web 2.0 technologies, a web-based system can aid medical researchers in conducting image-based Delphi studies for improved clinical guidelines and decision support.

Indiana University-Purdue University Indianapolis (IUPUI) / Biomechanical Big Data is of great significance to precision health applications, among which we take special interest in Physical Activity Detection (PAD). In this study, we have performed extensive research on deep learning-based PAD from biomechanical big data, focusing on the challenges raised by the need for real-time edge inference. First, considering there are many places we can place the motion sensors, we have thoroughly compared and analyzed the location difference in terms of deep learning-based PAD performance. We have further compared the difference among six sensor channels (3-axis accelerometer and 3-axis gyroscope). Second, we have selected the optimal sensor and the optimal sensor channel, which can not only provide sensor usage suggestions but also enable ultra-lowpower application on the edge. Third, we have investigated innovative methods to minimize the training effort of the deep learning model, leveraging the transfer learning strategy. More specifically, we propose to pre-train a transferable deep learning model using the data from other subjects and then fine-tune the model using limited data from the target-user. In such a way, we have found that, for single-channel case, the transfer learning can effectively increase the deep model performance even when the fine-tuning effort is very small. This research, demonstrated by comprehensive experimental evaluation, has shown the potential of ultra-low-power PAD with minimized sensor stream, and minimized training effort. / 2023-06-01

In order to arrive at informed judgments about the quality of research institutions and individual scholars, funding agencies, academic employers and researchers have turned to publication rankings. While such rankings, often based on journal citations, promise a more efficient and transparent funding allocation, individual researchers are at risk of showing adaptive behavior. This paper investigates whether the use of journal rankings in assessing the quality of scholarly research results in the unintended consequence of researchers adapting their research topics to the publishing interests of high-ranked journals. The introduction of the Handelsblatt Ranking (HBR) for economists in German language institutions serves as a quasi-natural experiment, allowing for an examination of research topic dynamics in economics via topic modeling and text classification. It is found that the Handelsblatt Ranking did not cause a significant shift of topics researched by German-affiliated authors in comparison to their international counterparts, even though topic convergence is apparent.

Scientific visualization is an essential and indispensable tool for the systematic study of computational (CFD) datasets. There are numerous methods currently used for the unwieldy task of processing and visualizing the characteristically large datasets. Feature extraction is one such technique and has become a significant means for enabling effective visualization. This thesis proposes different modules to refine the maps which are generated from a feature detection on a dataset. The specific example considered in this work is the vortical flow in a two-dimensional oceanographic dataset. This thesis focuses on performing feature extraction by detecting the features and processing the feature maps in three different modules, namely, denoising, segmenting and ranking. The denoising module exploits a wavelet-based multiresolution analysis (MRA). Although developed for two-dimensional datasets, these techniques are directly extendable to three-dimensional cases. A comparative study of the performance of Optimal Feature-Preserving (OFP) filters and non-OFP filters for denoising is presented. A computationally economical implementation for segmenting the feature maps as well as different algorithms for ranking the regions of interest (ROI's) are also discussed in this work.

No description available.