Human Computer Interaction Research Lab at SFSU
Computer Science Home | San Francisco State |CS Department Research Page  
HCI Lab Home
Research Overview
On-Going Projects
  Pen Computing
  Bioinformatics
  Information Visualization
  Smart Classroom
Publications
People
Projects for New Students
Collaborators and Funding

Pen Computing

In earlier joint work between Eric Saund at the Palo Alto Research Center and Edward Lank, Eric noted some of the advantages of pen-based input. Specifically, Eric noted that unlike the keyboard/mouse, the pen/stylus affords rapid and seamless mixtures of several kinds of interface actions:
  • input of open-ended symbolic information by writing text (keyboard)
  • input of open-ended symbolic information by drawing graphical symbols (hard to do with mouse)
  • unconstrained automatic positioning of text and symbols by pen/stylus placement during input
  • input/selection of commands from a limited command library by gesturing
  • selection of content by pointing and tapping (analagous to point and click with the mouse)
  • selection of content by spatial gestures such as encircling (sometimes difficult to do with a mouse)
  • navigation of and selection of items from menus (analagous to mouse)
  • direct manipulation of content (e.g. dragging to move, rotate, scale)
  • direct manipulation of widgets (sliders, buttons)
Unfortunately, pen-based interfaces have rarely achieved the fluid interaction that people expect from them. The result is that pen-based interfaces have yet to fully realize their potential as an interaction paradigm.

Our work in pen computing seeks to design more fluid interfaces based on a number of techniques. Specific projects we work on are briefly described here.

Inferred Mode Protocols
Inferred mode protocols examine user's actions and the context of those actions with the goal of partially eliminating modes in interfaces in pen systems. More information is available here.
Sloppy Selection
When users gesture with a stylus, they may gesture carefully or quickly. This project seeks to analyze the dynamics of pen motion to determine a tolerance for a selection gesture. More information is available here.
Focus+Context Sketching
Drawing on a small display presents its own challenges. This project seeks to design effective interfaces for use on small displays. To do this, we work on display surface distortion techniques to compress context information while providing a high-resolution focus area where a user can act. More details on this project are available here.
Sketch Recognition
Allowing users to enter information using a stylus is a challenging problem. Significant work exists in the area of handwriting recognition. As well, a number of researchers have designed input systems for pen-stylus interfaces geared toward a specific problem. Our work, however, seeks general solutions to the interpretation of hand drawn sketches and diagrams. Mor information on our sketch recognition work can be found here.