Shilman_parsing_aaai - Statistical Visual Language Models for Ink Parsing Michael Shilman Hanna Pasula Stuart Russell Richard Newton Department of

Info iconThis preview shows pages 1–2. Sign up to view the full content.

View Full Document Right Arrow Icon
Statistical Visual Language Models for Ink Parsing Michael Shilman, Hanna Pasula, Stuart Russell, Richard Newton Department of Computer Science University of California at Berkeley Berkeley, CA 94720 {michaels, pasula, russell, [email protected] Abstract 1 In this paper we motivate a new technique for automatic recognition of hand-sketched digital ink. By viewing sketched drawings as utterances in a visual language, sketch recognition can be posed as an ambiguous parsing problem. On this premise we have developed an algorithm for ink parsing that uses a statistical model to disambiguate. Under this formulation, writing a new recognizer for a visual language is as simple as writing a declarative grammar for the language, generating a model from the grammar, and training the model on drawing examples. We evaluate the speed and accuracy of this approach for the sample domain of the SILK visual language and report positive initial results. Introduction Since Ivan Sutherland pioneered pen-based computing with his SketchPad system over three decades ago (Sutherland 1963), there has been a widely-held vision of unencumbered tablet computers that present the feel of interactive, smart paper. Over years, we have seen numerous prototype systems that allow users express themselves directly in an appropriate syntax for different application domains, ranging from as flow-charts (Gross 1994) to mathematics (Matsakis 1999) to music notation (Blostein and Haken 1999). Even less structured domains like user interface and web page design can have their own domain-specific visual notation (Lin et al. 2000). Researchers have shown that such sketch-based applications can combine many of the benefits of paper-based sketching with current electronic tools to enable important new creative and collaborative usage scenarios (Landay and Myers 1995). Unfortunately, while we are on the verge of having suitable mass-market hardware devices to support the pen computing vision, we lack the software technology to adequately implement many of the most useful software applications that will run on these devices. This is not to say that researchers haven’t built a variety of toolkits to support sketch-based application prototyping. Existing toolkits support digital ink capture and storage, facilities for interpreting and beautifying sketched ink (Hong and Landay 00), and even sophisticated reusable schemes for user correction of incorrect interpretations (Mankoff, Hudson, and Abowd 2000). However, we believe that the problem of robust sketch recognition has been largely ignored and is crucial to the ultimate success of sketch- based user interfaces in the real world. The goal of this research is to move beyond prototyping and push recognition accuracies to a point where these systems are useful and predictable to end users.
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Image of page 2
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 06/12/2011 for the course CAP 6105 taught by Professor Lavoila during the Spring '09 term at University of Central Florida.

Page1 / 7

Shilman_parsing_aaai - Statistical Visual Language Models for Ink Parsing Michael Shilman Hanna Pasula Stuart Russell Richard Newton Department of

This preview shows document pages 1 - 2. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online