Expressive Spatial Interfaces for Scientific Visualization and Artistic 3D Modeling

This dissertation explores spatial human-computer interaction techniques to improve the control and expressiveness of 3D interactions. It investigates the requirements necessary for users to work more effectively with next-generation spatial interfaces, specifically in the context of scientific visualization and artistic 3D modeling where users currently struggle to express complex spatial concepts.;Examples of expressive spatial interfaces are presented and evaluated. In particular, we present new techniques for combining multi-touch with free-hand gestures for navigating visualizations and performing 3D surface modeling operations. Techniques for selecting and filtering volumetric data using lightweight props as well as active force-feedback are also introduced. Additionally, we present a spatial modeling interface for artistic 3D modeling using contextual interpretation of the user's input.;Several conclusions are drawn from these examples. Rich, parallel input and output streams enabled by recent advances in tracking hardware are particularly important for expressive interfaces. Additionally, there is a need for tighter integration of two and three-dimensional data and input. Contextual interpretation of user input enables users to specify more complex 3D concepts. Finally, many spatial tasks require immediate feedback to be expressive.;The primary contribution of this dissertation is a new class of interaction techniques called Expressive Spatial Interfaces that advance beyond the limited pointing and rotating interactions common in current-generation spatial interfaces. The techniques presented here can have a powerful impact on shaping the future of expressive spatial human-computer interaction with 3D graphics.