Skip to main content.

RATA.SSR

While many approaches to digital ink recognition have been proposed, most lack flexibility and adaptability to provide acceptable recognition rates across a variety of problem spaces. Time and expert knowledge are required to build accurate recognizers for a new domain. This project uses selected algorithms from a data mining toolkit and a large feature library, to compose a tailored software component (Rata.SSR) that enables single stroke recognizer generation from a few example diagrams.

RATA is available for download with instructions on how to generate and use your own recognizer.

RATA prototype
Screenshot of the RATA Recognizer Generator software

Publications

  • Chang S., B. Plimmer R. Blagojevic (2010). Rata.SSR: Data Mining for Pertinent Stroke Recognizers, 7th Eurographics Symposium on Sketch Based Interfaces and Modelling, Annecy, France, p 95-102.
  • Blagojevic R., S. Chang, B. Plimmer (2010), The Power of Automatic Feature Selection: Rubine on Steroids, 7th Eurographics Symposium on Sketch Based Interfaces and Modelling, Annecy, France, p 79-86.

    • This project was partially funded by Microsoft® Research Asia.