kurt

Morning Midterm Musings

In our meeting yesterday, Tripp and I discussed techniques for visualizing clusters of video in relational and geographic space. This issue overwhelmed our other significant challenge: selecting a meaningful analysis of the video. I am convinced that there are two major axes that must anchor our project: the temporal axis and the spatial axis. Without going into a long diatribe, my argument is that after the senses, temporal and spatial perception are the framework of our perspective on our environment. We like to see spatial and temporal representation of information, there is a deep satisfaction and great deal of insight gained through good visualization of data.

We do not (emphasis) analyze our visual experience in terms of brightness and contrast when we are constructing a mental mapping of our reality. Nevertheless, the capability of machines to provide us with data and metadata, extrasensory for humans, can be leveraged for projects where the additional information enriches our view of the environment.

My feeling is that the visualization of the video takes precedence and that the more mechanistic and abstract relationships between sequences will be the most meaningful when they are viewed in the context of space and time. So our napkin-based brainstorm produced a model of geographic space along the x,y that leaves a 3D map as a trail as it moves along the z axis. This is a complicated and dense visualization that needs to be seen and manipulated before it can be evaluated. The key would be in the creation of a smooth transition between this higher view and then subsequent closer views that drill down to the full-screen video itself.