国际学术会议模板.ppt

An open system model Globalization abc * * Research Challenges for Healthcare From Data to Knowledge to Action: Enabling Evidence-Based Healthcare Biomedical data, personal data (e.g., 24x7 activity and physiology traces), and relevant contextual data (e.g., indoor and outdoor air quality, exposures to toxins) Extraction, storage, and fusion of multiple classes of data from disparate sources and contexts * * Research Challenges for Healthcare Cont. Empowering people – providers and consumers – improves healthcare quality Human-computer interaction research The discourse between a patient and a doctor * * Research Challenges for Healthcare Cont. Computer-based knowledge system Human learning and reasoning Decision-making Physical motion significantly enhances human capabilities * * Research Challenges for Healthcare Cont. Healthcare is a complex, large-scale, adaptive distributed evolving system Researchers with different backgrounds and roles Policy and Investment Privacy and Security * * Evidence-Based Healthcare: Fall Detection * * Our System: HONEY HONEY: HOme healthcare sentiNEl sYstem HONEY: 3-steps detection * * Implementation FallSensor：HTC Smartphone Homeserver：Laptop，Microphone，Camera CloudPIS：Hadoop DFS * * Implementation（Cont.） * * -Determination of the Threshold of SVM Result 1: Distinguish between Fall and non-Fall accurately. -The accuracy of Speech Recognition： Result 2: Maintain high accuracy in quiet and noisy environments. Experiments Results * * Experiments Results（Cont.）Result 1 * * * * Experiments Results（Cont.）Result 2 * * Resolution File Size （MB） Upload Time(s) Download Time(s) Total Time(s) 640×480 97.4 2267 389 2656 352×288 40.6 945 162 1107 320×240 30.8 717 123 840 176×144 12.3 286 49 335 Experiments Results（Cont.）Result 3 Conclusion In order to improve the performance and accuracy, we deploy speech recognition and images to reduce the positive