Tim Lebo and Jon Matthis, RPI Graduate Students

 

Tim Lebo and Jon Matthis, RPI Graduate Students

Sage 4101

November 20, 2013 12:00 PM - 1:30 PM

Tim Lebo:

Title:                          Content-Preserving Graphics

Abstract:

Visual analytics is the science of analytical reasoning facilitated by visual interactive interfaces. As a discipline, it has advanced science and technology in analytical reasoning, interaction, data representations, visualization, and technology transition. Despite its successes, corporate practitioners of data analytics are still reporting many challenges. For example, in a recent survey of 35 analysts from 25 leading companies, Sean Kandel et al. report pervasive and costly difficulties in 1) discovering and wrangling data, 2) static results, 3) sharing of results, 4) discovering relevant data that becomes available, and 5) visualizations imposing a barrier to underlying data. Our hypothesis is that a dearth of explicit, meaningful connections within and among analytical results is a significant factor for the challenges currently facing practitioners of data analytics. We present a technique called "Content-Preserving Graphics" that maintains machine-understandable connections between the view of data and the underlying data itself. Content-Preserving Graphics is based on Linked Data design principles and reduces the amount of time for analysts to understand, trust, and re-purpose analytical reports created by previous, disparate, and perhaps unknown analysts. We will demonstrate the creation and use of a Content-Preserving Graphic using an application called the Data Sculptor, and finish by outlining future work to evaluate the impact of the Content-Preserving Graphics technique.

 

Jon Matthis:

Title:              The critical period for the visual control of foot placement occurs in the preceding step

Abstract:

In previous years, I presented results showing that  2-2.5 step lengths of visual look ahead is sufficient for humans to be able to walk over complex terrain while exploiting the basic biomechanics of bipedal gait as efficiently as they do with unlimited vision. In addition, last year I presented the results of a study showing that although subjects’ ability to step accurately to a target foothold was undiminished even when targets were rendered invisible for the entirety of the step towards them, stepping accuracy degraded rapidly if a target disappeared at any point during the step to the preceding target.  Taken together, the results of these two studies suggest that there is a critical period for the visual control of each step that occurs during the stance phase of the preceding step. In this talk, I present the results of a study designed to test this hypothesis by examining the behavior of subjects walking over a path of target footholds that were only visible during specific phases of their approach. Indeed, I found that each target only needed to be visible for a narrow window during the stance phase preceding the step in order for subjects to be able to place each foot accurately while efficiently conserving the inertia of their moving body. This simple control strategy allows humans to accurately walk over indefinite stretches of complex terrain while efficiently exploiting the biomechanics of human gait.

Add to calendar
Share|