Quality-Adaptive Media-Flow Architectures to Support Sensor Data Management

该项目开发了新的质量自适应实时媒体流体系结构，以支持感觉/反应环境。亚利桑那州立大学的艺术、媒体和工程(AME)中心拥有一个先进的表演舞台，被称为智能舞台，允许表演者通过各种传感器和执行器与他们的环境进行新颖的互动。在这个框架内，我们为互动艺术(ARIA)开发了一个自适应和可编程的媒体流架构，使互动表演的设计、模拟和执行成为可能。ARIA与实时传感组件接口，它可以传输各种类型的音频、视频和运动数据。它从流数据中提取各种特征，并根据服务质量(Qos)要求将媒体流融合和映射到输出设备。ARIA为编导提供了一个可视化的设计工具来描述媒体流网络的结构，其中的节点对应于传感器、质量自适应计算元素和执行器。我们开发了新的媒体流管理技术，包括实时和服务质量驱动的媒体感知、特征提取以及数据和媒体融合算法。该项目的影响包括更好地了解基于服务质量的数据和媒体流管理。ARIA提供了一种新的媒介，允许艺术家在舞台表演中整合新的感知、互动、内容和反应机制。ARIA还充当了一个研究/测试/开发的平台，工程和艺术专业的学生可以在这里学习和试验多媒体流管理的各个方面。该项目所取得的成果将有助于更好地理解以质量为基础的数据和流管理问题，因此将影响新型传感器数据管理技术的发展。项目网站是http://aria.asu.edu.ARIA，正在通过亚利桑那州立大学的艺术、媒体和工程中心向公众传播。

This project develops novel quality-adaptive real-time media-flow architecture to support sensory/reactive environments. Arts, Media, and Engineering (AME) Center at Arizona State University houses an advanced performance stage, known as the Intelligent Stage that allows performers to have novel interactions with their environment through various sensors and actuators. Within this framework, we develop an adaptive and programmable media-flow ARchitecture for Interactive Arts (ARIA) that enables design, simulation, and execution of interactive performances. ARIA interfaces with real-time sensing components and it streams various types of audio, video, and motion data. It extracts various features from streamed data, and fuses and maps media streams onto output devices as constrained by the Quality of Service (QoS) requirements. ARIA provides the choreographer with a visual design tool to describe the structure of the media-flow network, where the nodes correspond to sensors, quality-adaptive computing elements, and actuators. We develop novel media-flow management techniques including real-time and QoS driven media sensing, feature extraction, and data and media fusion algorithms. The impact of the project includes a better understanding of QoS-based data and media-stream management. ARIA offers a new medium that allows artists to integrate novel sensing, interaction, content, and response mechanisms in staged performances. ARIA also acts as a research-/test-/development-bed where engineering and arts students learn and experiment with various aspects of multimedia stream management. The results obtained by this project will lead to a better understanding of quality-based data and stream management issues and will therefore influence the development of novel sensor data management technologies. The project web site is http://aria.asu.edu.ARIA is being disseminated to the public through the Arts, Media, and Engineering Center at ASU.