Projects

Ongoing
	NSF Research Experience for Undergraduates: Smart Cities 10 Week summer research program for undergraduate students from across the country to participate in guided research in intelligent transportation systems (ITS), connected automated vehicles (CAV), and vehicle-to-everything communication (V2X).
	Connected Automated Vehicle (CAV) We are developing a CAV program at UNLV using the Autoware open source autonomy stack. Emphasis on is environmental perception and integration with infrastructure.
	NSF EPSCoR Fire Science Nevada project to increase the capacity for wildfire research, education, and workforce development. Our effort is focused on machine learning for fire process modeling.
	Tissue Perfusion Joint work with surgeons in the UNLV Medical School to develop a non-contact device to monitor patients post surgery for tissue perfusion.
	Action Quality Assessment Use of deep learning to automatically characterize how well an action was performed. Experiments performed using Olympic sports.
	Deep Trajectory Prediction Use of deep learning to accurately predict a long-term trajectory. Research considers how to leverage individual dynamics, social interactions, and scene information to predict pedestrian trajectory four seconds into the future. The goal is to have models of behavior which can be used in automated vehicles.
	MySnapFoodLog: Camera Food Tracking Develop a camera food logging app to study the affects of dietary biculturalism (how does diet change for immigrants in the US) in Filipino population. Collaboration with Nursing and Nutrition. Android App
	TUG Calc: Timed-Up and Go Calculator Develop an app to compare TUG time versus groups with similar characteristics such as age, bmi, or medical conditions. Provide more fine-grained analysis of risk of fall compared with basic TUG recommendation. Collaboration with Physical Therapy. Android App
Completed
	Maritime Port Activity Monitoring Comprehensive monitoring of maritime environments. Use of fine-grained classification to distinguish different types of vessels. Tracking information used to build contextual scene description for understanding of activities and detection of anomalous events.
	Accurate predictions of PV power production on small-time scales Development of image processing and computer vision algorithms to detect, measure, and predict cloud movement patterns which can occlude the sun and lower PV output. Using cloud image-based measurements, develop a prediction algorithm to map the observed sky characteristics to PV output in the near future by exploiting convolutional neural networks and deep learning.
	Intersection Monitoring Detection and tracking of intersection road users for improved transportation systems. Automatically calculate intersection turn counts for operations and examine trajectories for safety.
	Kinect Rehabilitation Games Develop gaming environment to promote at home physical therapy (PT) in children with mobility disorders. The Microsoft Kinect provides a low cost alternative to traditional physical therapy treatments with the convenience of at home use. The serious gaming environment should encourage PT exercises in an enjoyable manner.
	Next Generation Performance Monitoring Data Needs for Nevada DOT Develop a comprehensive performance measurement plan for Nevada. Provide an inventory of state-wide data collection, identify gaps in data, and provide summary of applicable best-practice performance measures. Sponsor: Nevada Department of Transportation Date: 2014
	Nevada Regional License Plate Recognition Characterize license plate recognition (LPR) technology for Nevada and southwest regional license plates. Design activity awareness and analysis for a garage-based LPR installation. Sponsor: Image Sensing Systems, Inc. Date: 2014