Making sign languages easier to study and process with computers

What is it about?

Sign languages are visual languages that use hand movements, facial expressions, and body posture to communicate meaning. Because they do not have a widely adopted writing system, we need reliable ways to describe and annotate them so that both people and computers can understand the data. This study proposes an annotation method for Italian Sign Language (LIS) that combines ideas from sign language linguistics with the practical needs of artificial intelligence and automatic sign recognition. Instead of relying only on traditional glosses (written words used to represent signs), the method combines written labels in both Italian and English with a language-specific annotation system called Typannot. This makes the data more informative while keeping it readable for researchers, signers, and computer scientists.

Featured Image

Photo by Flavie Martin on Unsplash

Photo by Flavie Martin on Unsplash

Why is it important?

As sign language datasets become increasingly important for both linguistic research and artificial intelligence, there is a growing need for annotation methods that accurately represent how sign languages work while remaining suitable for computational processing. The methodology proposed in this article can support the creation of higher-quality sign language datasets, facilitate interdisciplinary collaboration, and contribute to the development of more accurate automatic sign recognition systems and future language technologies for sign languages.

Perspectives