Construction of a System for
Recognizing Touch of Strings for Guitar
Hayami Tobise
Kobe University
Kobe, Japan
tobise@stu.kobe-u.ac.jp
Yoshinari Takegawa
Future University Hakodate
Hakodate, Japan
yoshi@fun.ac.jp
Tsutomu Terada
Kobe University/PESTO, JST
Kobe, Japan
tsutomu@eedept.kobe-
u.ac.jp
Masahiko Tsukamoto
Kobe University
Kobe, Japan
tuka@kobe-u.ac.jp
ABSTRACT
In guitar performance, fingering is an important factor. In
particular, the fingering of the left hand comprises various
relationships between the finger and the string, such as the
finger touching/pressing/releasing the strings. The recog-
nition of the precise fingering is applied to a self-learning
supp ort system, which is able to detect strings being muted
by a finger, and which transcribes music automatically, in-
cluding the details of fingering techniques. Therefore, the
goal of our study is to construct a system for recognizing
the touch of strings for the guitar. We propose a method
for recognizing the touch of strings based on the conductive
characteristics of strings and frets. We develop a proto-
type system, and evaluate its effectiveness. Furthermore,
we propose an application that utilizes our system.
Keywords
Guitar, Touched strings, Fingering recognition
1. INTRODUCTION
Fingering is an important aspect of guitar performances.
In guitar performances, fingering affects the ease of play-
ing a phrase as well as musical expression such as timbre.
Guitarists explore the best fingerings to best convey their
expressions, so there are many styles of fingering, which
are different for each guitarist. The real-time recognition
and distinction of the fingering of the left hand is applied
to a self-learning support system, which is able to detect
incorrect fingering. Our research group has developed the
real-time fingering detection system for contrabass by inte-
grating simple camera-based image processing and musical
rules [1]. However, the system only recognizes the position
of the guitarist’s left fingers, when he/she plucks the strings
of the contrabass.
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that copies
bear this notice and the full citation on the first page. To copy otherwise, to
republish, to post on servers or to redistribute to lists, requires prior specific
permission and/or a fee.
NIME’13, May 27 – 30, 2013, KAIST, Daejeon, Korea.
Copyright 2013, Copyright remains with the author(s).
Figure 1: Relationships between finger and string
In guitar performances, the fingering of the left hand
comprises various relationships between the finger and the
string, such as a finger touching/pressing/releasing the strings,
as shown in Figure 1. The recognition of precise fingering of
the left hand contributes to a self-learning support system,
which detects strings being muted by a finger and touched
by a finger before being pressed, and which transcribes mu-
sic automatically with the details of fingering techniques.
The goal of our study is to construct a system for rec-
ognizing the touch of strings for the guitar. The proposed
system has a mechanism for recognizing the touch of strings
based on the conductive characteristics of the guitar’s strings
and frets. This method does not require any modification
on the guitar itself, simply by attaching the proposed de-
vice, which does not interfere with guitar performances.
We define the three relationships between the fingers and
the strings, and we implement an application utilizing our
metho d.
The remainder of this paper is organized as follows: Sec-
tion 2 describes the design of the recognition system, Sec-
tion 3 explains our evaluation and discusses the results, Sec-
tion 4 describes the implementation of applications, Section
5 explains related work, and finally Section 6 describes our
conclusions and future work.
2. DESIGN
There are two requirements for constructing our fingering
detection system:
(1) No interference with guitar performances
(2) Accurate and high-speed recognition
(1) means that it is supposed to be used in situations where
a player has to concentrate, such as in a concert, a les-
son, or self-teaching. Therefore, the performer should not