Photo: Unsplash.

Abstract

Analysis of negative headshake in sign languages using Computer Vision: first results

Nonmanual markers play a crucial role in sign language grammar. For instance, in many sign languages, headshake is used to express negation: either in combination with manual signs, or even on its own.

Previous studies have found differences between sign languages in frequency and co-occurrence patterns of negative headshake; however, the phonetic/kinetic properties of headshake have not been explored due to methodological constraints.

Recent developments in Computer Vision have made it possible to analyze sign language data, including nonmanuals, using a quantitative approach. For example, it is possible to use Head Pose Estimation to measure rotation of the head in video recordings of sign languages, and thus use these measurements to quantify and analyze linguistically relevant movements, such as headshakes expressing negation.

In my presentation, I will share the first results of an ongoing exploration of phonetic properties of negative headshakes crosslinguistically, focusing on three sign languages: NGT, DGS, and Libras. Using naturalistic corpus data, I will show that linguistic factors, such as the presence or absence of spreading and a manual negative sign, can affect phonetic properties of headshakes. 

On Professor Vadim Kimmelman