DIAGRAM Subcontract Ideas for 2013

Skip to end of metadata
Go to start of metadata
Ideas should promote the 2012-2015 DIAGRAM Goals
  • Reference implementation of tools supporting accessible math, multi-modal accessible images, and content model so tools builders and content creators know what works
  • Support for interactive accessible image widgets so we can inspire more smart image creation
  • Enhance accessible math navigation and explore math manipulation and annotation so students can work better with math content 
  • Accessible Image Coverage checker for EPUB3 so content creators can tell if they are creating accessible materials
Potential Ideas
  1. Tools for tagging resources that have long image descriptions and MathML, see http://a11ymetadata.org
  2. MathJax aural rendering
  3. Tools for crowdsourcing of image descriptions for the web based resources (similar to old Webvisum). For example:
    1. Tools for WordPress authors/publishers to easily create accessible infographics via a WordPress plugin.
    2. BVI readers can request that inaccessible graphics be made accessible by using a browser extension that posts comments to the author and other readers with a link to a tool to make the offending image accessible.
    3. In cases where the author or publisher does not correct the accessibility issue at the source or does not have the WordPress plugin, the same browser extension enables readers to access cloud hosted alternative accessible versions created by other readers/volunteers via the browser extension.
  4. Tools for promoting use of longdesc and use of the DIAGRAM Content Model on the web, particularly blogs and online learning resources.
  5. Tools for creating accessible image widgets: collect and gather, hackathon?
  6. Tools such as the GLANCES project proposed by SKERI for image classification.  See Glances Project below.
  7. Conversion of 2D SVG graphics into files that can printed with 3D printers.  The 3D printing market is continuing to grow and solutions are arising for print on demand when you don't have your own printer.  An end-to-end solution involving Poet for creating/attaching SVG alternatives that have tour elements for audio description using the content model would be fantastic.  This can tie to the item below and Lore's idea of CAD skilled students volunteering to create 3D / SVG alternatives:
    1. https://www.dropbox.com/s/vbm9127ds0f8xw6/3D%20to%20Image%20File%20Crowd-Source.doc?n=62631409
  8. Using increasingly cheaper gesture recognition technology, such Leap Motion ($80) and Microsoft Kinect, to track a BVI user's hand and fingers as they explore a tactile graphic and provide an audio description/tour. (Bruce: Leap doesn't work well for tracking fingers on a touch screen)
  9. Building sonification into Benetech web based version of MathTrax through use of the HTML5 Web Audio API or other simpler mechanisms that leverage the existing server generation of the audio files.  It would be good to see this web-based MathTrax linked to the sample book project.
  10. Pushing forward haptic work that Mark Hakkinen is conducting at ETS (Geoff, John G)
  11. Polymers ? (George)
  12. Accessible image support in Readium web reader and/or Readium SDK

We envision a tool that can categorize the type of the diagram, and automatically extract simpler information. SKERI has already done some preliminary work on a diagram categorization tool under the RERC grant, and we will build on this to provide a much more extensive tool. The main tool may would automatically decide on the type of diagram, such as pie chart, or flowchart or state diagram, and pass this on to other software specifically tailored to deal with such diagrams. Then, the secondary tool, for example a pie-chart describer, may determine the slices of a pie chart, determine the corresponding annotations in the figure, order them in order of decreasing size, and present the results in tabular form. A human annotator can then proofread the results and fix any errors. If the system works as well as we hope, this may lead to significant time/cost savings for a tool such as Poet. We are also hoping to release this tool as a standalone open-source project.

Enter labels to add to this page:
Please wait 
Looking for a label? Just start typing.