CAREER: Managing Complexity: Fidelity Control for Optimal Usability in 3D Graphics Systems

随着3D图形渲染硬件速度的提高，显示模型的尺寸也大幅增加。现在，全系统CAD、3D模型扫描仪、程序生成模型和可视化数据集的大小接近数亿甚至数十亿个多边形。试图显示这些模型的研究人员被迫要么降低显示速度和交互性，要么降低模型显示视图的保真度。当模型复杂性自动降低或“简化”时，保持视觉保真度的最佳方法是什么？在显示速度和视觉保真度之间达成妥协的最有效方法是什么？在日益重要的可视化图形应用中，我们如何保留显示数据中有意义的元素，并使用户成为信息复杂性的有效管理者？本研究将从以上三个问题出发，采取三个互补的研究方向。PI将开发原型系统，并通过用户研究来调查其有效性。这些研究中有许多都是在与学术界和游戏行业的同事合作开发应用程序的背景下进行的。具体目标将包括开发和评估测量保真度的自动方法；检验保真度控制应该最小化视觉误差的假设，无论它是由视觉保真度或显示速度的妥协引入的；对于抽象信息可视化，保真度测量和总结技术的研究不一定受到自然世界感知知识的启发。虽然许多技术解决方案已经解决了测量相似性和汇总复杂数据的问题，但其中只有少数在严格的可用性研究中得到了评估。这项研究将填补这一空白，产生知识，指导方针，启发和软件，将提高交互式3D图形应用程序的可用性。

Improvements in the speed of 3D graphics rendering hardware have been accompanied by even more drastic increases in the size of displayed models. Full-system CAD, 3D model scanners, procedurally generated models, and visualization data sets are now approaching hundreds of millions or even billions of polygons in size. Researchers trying to display these models have been forced either to reduce display speed and interactivity, or instead to reduce the fidelity of the displayed views of their models. What are the best methods for preserving visual fidelity as model complexity is automatically reduced or "simplified"? What is the most effective way of striking the display speed vs. visual fidelity compromise? And in the increasingly important graphics application of visualization, how can we preserve the meaningful elements of the displayed data, and make users effective managers of information complexity? This research will take its three complementary directions from the three above questions. The PI will develop prototype systems and investigate their effectiveness with user studies. Many of these studies will take place in the context of driving applications developed for existing collaborations with colleagues in academia and the gaming industry. Specific goals will include the development and evaluation of automatic methods for measuring fidelity; an examination of the hypothesis that fidelity control should minimize visual error, whether it is introduced by compromises in visual fidelity or display speed; and for abstract information visualization, study of fidelity measures and summarization techniques not necessarily inspired by knowledge of perception in the natural world. While many technical solutions have addressed the issues of measuring similarity and summarizing complex data, only a few of them have been evaluated in rigorous usability studies. This research will fill this gap, producing knowledge, guidelines, heuristics and software that will improve the usability of interactive 3D graphics applications.