Menu Close

Off-Screen Desktop


Erik Paluka and Christopher Collins

We present Off-Screen Desktop, spatial navigation techniques that make use of the space around the display to extend direct manipulation beyond the desktop screen. To enable off-screen direct manipulation, these techniques visually transform the information space without affecting its interaction space. This allows a person to interact with the information space as if it physically extended beyond the boundaries of the display. Off-Screen Desktop is characterized by its implicit transience where the applied visual transformations are automatically reverted when the hand leaves the associated spatial interaction space. We illustrate Off-Screen Desktop with the design of three different techniques, which include Dynamic Distortion, Spatial Panning, and Dynamic Peephole Inset, as well as their evaluation in a comparative study with standard mouse panning. We also demonstrate their applicability with a number of use cases. Study results show that Spatial Panning was overall significantly faster than the other Off-Screen Desktop techniques when employed in two different navigation tasks.

Check out our GitHub Repository for source code related to this project.


    [pods name="publication" id="4326" template="Publication Template (list item)" shortcodes=1]



Dynamic graphs, actively used domains such as social, biological, or computer network analysis, are challenging to visualize and explore due to simultaneous topological changes occurring over time. Glidgets presents a new combined direct manipulation and visualization technique for exploring and querying dynamic graphs. Traditional approaches provide an indirect time slider and employ visual cues such as global change highlighting. This work merges temporal navigation and the visual representation of graph dynamics into new interactive visual glyphs on nodes and edges. Interactive timeline glyphs reveal the presence and absence of nodes and edges, and node degree. Using sketch-based interaction, the glyphs are used to create queries and navigate time directly on graph nodes and edges. This enables one-stroke gestures to create queries such as “Are these nodes ever connected?” or “When is this node present in the network?” Analysts can directly query changing graph elements and investigate those changes by navigating time while remaining focused on the element of interest.

Try our demo!


    [pods name="publication" id="4335" template="Publication Template (list item)" shortcodes=1]