Creating and Evaluating New Media Interfaces for Expression

点击翻译按钮获取中文摘要

I propose to push the boundaries of two forms of interactive visual media to increase the capacity for humans to learn and work together. First, new forms of interaction with a novel 3D display will be investigated. Second, advanced interactive, collaborative video (IV) with three critical new affordances to elevate video to be a `first-class' learning medium will be created. First, research using our 3D fish-tank virtual reality (FTVR) spherical display (3DD) focuses on two main types of interactions: 1) co-located and remote multi-person interaction requiring real-time 3D capture, and 2) human-3D agent interaction. To support co-located four-person interaction, a curved polarizing lens will be created to match the curvature of the spherical display surface coupled to active glasses to enable four different monoscopic perspective corrected views to be projected on the display but seen by different users. These displays provide virtual contents anchored to real-world coordinates allowing everyone to share a common reference frame. Thus, deictic interactions, like pointing and saying "That thing there." are supported where everyone can see what is being pointed at. For remote participants, a real-time 3D capture system using depth cameras supporting eye-contact will be created. Perceptual effects as well as task performance on 3D tasks will be studied and compared to similar configurations using flat screens. Second, viewer and room interaction with an intelligent virtual agent (IVA) inside the spherical display will be studied. As with real-participants, voice and gesture interaction will investigated. Likewise, interaction semantics for collaborative pointing, selecting, translating, and rotating content will be explored in remote and co-located scenarios. This research will create advanced 3D spherical displays for groups of people and agents to work and play across space and time. Second, from past work, we discovered three video affordances critical for education are missing: mutability, bidirectionality and interoperability. For mutability, interfaces for collaborative video making while each state change and annotation is managed will be created. For interoperability, interfaces supporting video structure and semantic capture and copy for effective integration with students' information ecology, including notes will be created. For bidirectionality, rich referencing interfaces enabling complex methods of communicating with video elements will be investigated. These interactions will be studied in educational scenarios such as: note taking, presenting, and social learning. Ultimately, the goal of this research provides the `pen and paper' for video so people can learn to `read/watch and write/compose' video as easily as they can through text. This research directly involves 33 highly qualified personnel working in world-class international, interdisciplinary teams to become leaders in advanced media interaction.