a Tool for Synthesizing
Animation of Sign Language

by Miki Grahame



Winner of the BTexact Technologies award for Innovation in Computer Graphics

Animation produced with HandsOn
HandsOn Explained
Project Summary



HandsOn was a 9-week project produced during my master's degree course.

My aim was to develop a tool for creating computer-generated animation of sign language. Building on research findings of others, I implemented a working prototype, which synthesised animation of sign language hand gestures. The tool was designed to handle a general class of manual gestures, entered in a custom-designed gesture description language.

Looking at recent work, it seemed that many projects of this kind were not taking advantage of the sophisticated animation software now available. So I geared my tool to work in conjunction with Maya to create production-quality animation.


Project Summary

Sign Language involves a wide range of hand gestures and is therefore difficult to simulate in CG. A prototype tool, HandsOn, was developed to simplify the process of generating such animation without using motion capture. The tool takes as its input a textual description of a gesture sequence, coded in a specially designed language. Control systems are used to give the motion a more life-like quality. The program, written in C++, generates the animation and renders a simple visualisation of the sign gestures in real-time using OpenGL. The animation key-frames can be ported into a skinned Maya model in order to be rendered at high resolution.

The animation produced using HandsOn showed visible improvement over other existing sign language synthesis programs. With further work the generated animation could ultimately look as authentic as motion capture. There is also great potential for using the technique for other animation applications.


Programming Languages

If you'd like to try HandsOn, email me for a copy of the source code




Miles Green - modelling and skinning of the Maya hand models

Cazz O'Neill - sign language translation

Thanks to:

Richard Kennaway, who is developing a sign language synthesis system as part of the VisiCAST project

NCCA staff and students: Ari Sarafopoulos, Sami Nikki, Adam Vanner, Ben Morgan, Andy Hayes