Foldable Interactive Displays

Related comments:


Summary:
Authors Johnny Chung Lee, Scott E. Hudson, and Edward Tse (from Carnegie Mellon's HCI Institute and Smart Technologies respectively) discuss their research on foldable displays in this paper.  Most "flexible display" prototypes that are typically reported about are based on OLED (organic light-emitting diode) technology, which are still limited in their implementation and offer no form of touch input.  The foldable displays discussed by the authors are actually flexible screens with embedded infrared (IR) LEDs that allow a camera/projector combination to track the display and project an appropriately-sized image onto it.  The IR LEDs basically function as fiducial markers in an augmented-reality system where digital images (in this case, whatever is supposed to show up on the display) are superimposed over real life.  The authors discussed the advantages and drawbacks of four main foldable shapes -- the newspaper, scroll, fan, and umbrella -- as well as how IR tracking is accomplished on each shape.

Discussion:
When I read the title of the paper, I assumed that the displays would be driven by OLED and that the authors had somehow solved the problem of touch input on such a display.  Alas, it was not to be as they basically rigged an augmented-reality system together with IR LEDs and a Wii-mote which allowed camera-based tracking and projection onto the static, un-interactive surface.  The displays they discussed could have been made of anything -- paper, cloth, plastic, whatever.  The whole system is driven by the camera tracking.  I suppose I shouldn't have gotten my hopes up so high, but it's difficult to see how this system could be used anywhere outside of a classroom or research facility with the necessary technologies already installed.  The point of smaller, mobile displays is to be able to carry them with you without the need for external equipment.  The only real advantage I can see here (other than the orientation-sensitivity, which was kind of cool) is that a user would be able to carry a movie-screen-sized display in a backpack or even their pocket and set it up with minimal effort, assuming that the IR-tracking and camera equipment is available wherever they planned on watching a movie.

What's Next?: Emergent Storytelling from Video Collections

Related comments: no other posts on this paper, comment on Kerry's blog.


Summary: 
Three researchers from MIT's Media Laboratory (Edward Yu-Te Shen, Henry Lieberman and Glorianna Davenport) developed Storied Navigation, a video editing system that lets users quickly search and browse a video library to compose a story.  The system works by annotating video clips with sequence information and story attributes; users can then search with words or sentences to find clips with relevant characters, themes, emotions and story elements.  In this way, users simply type in an emotion or story sequence they wish to convey and the program returns clips matching their criteria -- the user doesn't even have to be familiar with the video library to be able to tell a story.  Storied Navigation uses natural language  processing (NLP) to parse user input for story elements, characters, etc.  It features an "edit-by-typing" function that sequences video clips based on a story the user types in English, and an "edit-by-recommendation" feature that can either find similar alternatives to selected clips or provide clips that will continue the current story sequence ("what's next").


The researchers conducted two separate studies, the first of which focused on whether the system helped users to develop their story threads.  The subjects reported the system's ease of use, helpful features, powerful search, the improved efficiency with which they could sift through large amounts of video to find what they needed, ability to compose stories on-the-fly, and the way the interface helped them understand the logic behind a story.  The second study focused on browsing an unknown video library.  All seven subjects reported that the system helps them find what they want, mentioning that the edit-by-typing function was the most useful to them -- some of the subjects even wanted to create stories from the clips that were returned from their queries.


Discussion:
BRILLIANT.  As a video editor, I can tell you that there is nothing more painful than trying to sift through hours of documentary footage to find an appropriate clip -- this tool is literally the coolest thing a documentary filmmaker could ever want (aside from one of these, maybe).  There are times during editing when I even know what all of the footage is (in the event that it isn't dozens of hours of material), and just don't know how to move forward with it.  The Storied Navigation system answers my unspoken question ("what's next?") and even provides multiple ways of filtering the results it provides by matching emotions or placement in the story structure.  The ability to compose stories on-the-fly or quickly change a train-of-thought to another direction greatly streamlines the film editing process.  I could save so much time and tedious browsing by using this system; the "ways to think" that it provides are simply brilliant and ought to spur a storytelling sense in any filmmaker.