Information Technology Reference
In-Depth Information
Figure 2. Illustration of the lazy interpretation
process
Figure 3. Illustration of the eager interpretation
process
requested by the user) or eager ( on the fly, , that is
occurring directly while the user is drawing). As
presented on Figure 2, lazy interpretation offers
the advantage of not intruding into the user's
creative phase during the composition because he
can write everything he has in mind before asking
the system to interpret the document (Nakagawa,
Machii, Kato, & Souya, 1993). Nevertheless, once
the recognition process applied at once on all of
his strokes, he has to examine the entire docu-
ment to look for possible incorrect interpretations.
Besides, it turns out that lazy systems are so far
not robust enough and make too many mistakes,
which reduces their usability. We believe that
lazy recognition is a promising approach to offer
unconstrained understanding of ink, but the dif-
ficulties to design automatic parsing coupled with
a robust hand-drawn shape recognition system
show that it remains an open problem.
Eager interpretation is then another way to
consider online structured document analysis. As
presented on Figure 3, every time the user draws
a stroke, the system interprets it immediately
(Blostein, Lank, Rose, & Zanibbi, 2002; Macé,
Anquetil, & Coüasnon, 2005a); it then has to deal
with documents directly during their composition
and to make a decision as quickly as possible so
that the user does not have to wait to continue his
drawing. Eager interpretation allows exploiting
the interaction with the user, who is aware of the
system answers, and can then validate or refute
them progressively. We believe that eager inter-
pretation is the most pertinent compromise for
online structured document interpretation because
systems based on this kind of interaction are often
more robust and more efficient than lazy ones,
which makes them more usable. The work we
present aims at exploiting a generic approach in
order to eagerly interpret various forms of hand-
drawn musical notations and various composition
conventions.
Whereas the existing approaches only enable
users to draw classical musical score notations,
our approach aims at making it possible to easily
adapt the system to different ones. In this chapter,
we focus on the composition and the editing of
various musical notations, such as classical musi-
cal scores, plainchant scores, drum tablatures and
stringed-instrument tablatures, but the method is
generic and has already been exploited for two
others structured document domains, which are
diagrams and unified modelling language (UML)
class diagrams.
In the following section, we present a state
of the art in pen-based interface development in
the domain of musical notation composition and
editing. We first state that, so far, systems are dedi-
cated to classical musical scores. We emphasize,
on the one hand, their differences with traditional
composition conventions, and, on the other hand,
their limitations in term of musical notations. In
section “A Generic Approach for Hand-Drawn
Structured Document Composition and Editing”,
we introduce our generic approach and present
each of its components. In section “Pen-Based
Systems for Music Composition and Editing with
the Presented Approach”, we present how this
generic approach can be used to compose and
edit four different natures of musical structured
documents: classical musical scores, plainchant
scores, drum tablatures and stringed-instrument
tablatures. Finally, we conclude this chapter by
highlighting future trends in the exploitation
of pen-based interaction for musical notation
composition and editing and by presenting our
future works.
Search WWH ::




Custom Search