High volume of Full Motion Videos (FMVs) require development of automated tools to help reduce the cognitive burden of the analysts. The number of algorithms for object detection, classification, tracking and anomaly detection in FMV were investigated. The object detection and classification was performed using YOLOv4 technique. Two approaches for object tracking were analyzed: (i) short-term tracking approach based on DeepSORT and (ii) persistent tracking based on template matching and structural similarity index. Anomaly detection and pattern of life analysis algorithms based on trajectory clustering and time series analysis were tested on simulated data and real FMV over a highway with traffic.

General health examination is an integral part of healthcare in many countries. Identifying the participants at risk is important for early warning and preventive intervention. The fundamental challenge of learning a classification model for risk prediction lies in the unlabeled data that constitutes the majority of the collected dataset. Particularly, the unlabeled data describes the participants in health examinations whose health conditions can vary greatly from healthy to very-ill. There is no ground truth for differentiating their states of health. In this paper, we propose a graph-based, semi-supervised learning algorithm called SHG-Health (Semi-supervised Heterogeneous Graph on Health) for risk predictions to classify a progressively developing situation with the majority of the data unlabeled. An efficient iterative algorithm is designed and the proof of convergence is given. Extensive experiments based on both real health examination datasets and synthetic datasets are performed to show the effectiveness and efficiency of our method.

Reports on research to adapt the 2/sup nd/ Generation Prairie Agrometeorological Model for wheat developed by Raddatz (1993) at Environment Canada to use both in situ data from a distributed network of meteorological stations (original mode) and NOAA AVHRR information output from the GeoComp-N processor at the Manitoba Remote Sensing Centre. Two parameters previously approximated by the model have been substituted with AVHRR derived information: fractional leaf area, and skin temperature. These parameters are critical for obtaining accurate outputs. Model outputs were ingested into a GIS to generate maps of crop growth stage expressed as biometeorological time, actual evapotranspiration or crop water-use, and top zone and root zone soil moisture for a test area in southern Manitoba for the year 2000 growing season.

We describe the role of mentors in en experimental murse on human-computer interaction (HCI) taught in the Computer Science Department at Stanford University for the past two years.Students practice design within the course by collaborating in smell groups on 12-week projects, in which they analyze a work environment, design and implement a prototype user interface, and evtthtate the prototype with project clients.As part of our experiment in providing contact with the world of practical design, we invited people from local industry to serve es mentors for the stmdent groups.These unpaid volunteers have been important adjuncts in guiding students as they developed their HCI projects.We dkeuss: the background for the projects; the role of mentors in the learning process; what is required of mentors end what benefits they ge~ how to support the mentorirtg process; end what we have learned horn the experience so far.We consider mentors to be a critical pert of the design project within the course, end we welcome a sharing of experience with other people who may have developed similar courses where mentors played an important role.1.