T. Gross et al. (Eds.): INTERACT 2009, Part I, LNCS 5726, pp. 196–199, 2009.
© IFIP International Federation for Information Processing 2009
Automatic Translation System to Spanish Sign
Language with a Virtual Interpreter
Sandra Baldassarri, Eva Cerezo, and Francisco Royo-Santas
Advanced Computer Graphics Group (GIGA), Computer Science Department,
Aragon Insittute for Engineering Research (I3A), University of Zaragoza, Spain
{sandra,ecerezo,497328}@unizar.es
Abstract. In this paper, an automatic translation system from Spanish language
into Spanish Sign Language (LSE) performed by a virtual interpreter is pre-
sented. The translator is based on rules from Spanish grammar considering the
syntactical and morphological characteristics of words and the semantics of
their meaning. The system has been incorporated to an animation engine in
which a virtual character acts as an interpreter that communicates using LSE.
The mood of the interpreter is considered, so that the translation and the signs
are modified depending on whether the interpreter is happy, angry, etc. The sys-
tem has been tested with satisfactory results in speed and quality.
Keywords: Sign language translation, virtual characters, animation, emotion.
1 Introduction
In the last few years, the design of computer application interfaces has evolved in
order to guarantee the accessibility of applications to everyone. Regarding the deaf
community, a considerable amount of work has been done in the automatic translation
into sign languages (SL). These languages, unfortunately, are not universal and each
country has its own variety. In fact, most of the work done [1] is based on English
grammar. This is the case of the works derived from ViSiCAST [2] and eSIGN [3]
projects. Regarding Spanish Sign Language (LSE), San-Segundo et al. in [4] have
developed a translator based on Spanish grammar that uses VGuido, an eSIGN avatar,
but their application domain is very restricted (sentences spoken by an official when
assisting people who are applying for their Identity Card).
None of the previous works includes mood. As in face-to-face communication,
mood, emotions and facial expressions are an integral part of sign languages [5].
Words can considerable change their meaning depending on the mood/emotion of the
speaker. Moreover, communicating in sign language without facial expressions would
be like speaking in a monotonic voice: more boring, less expressive and, in some
cases, ambiguous. The system presented in this paper is based on Spanish grammar,
takes mood into account and has been integrated in an existing animation engine ca-
pable of managing expressive virtual characters to perform the signs.