Artificial Intelligence for Sign Language Recognition and Translation
Master thesis
Permanent lenke
https://hdl.handle.net/11250/3020364Utgivelsesdato
2022Metadata
Vis full innførselSamlinger
Sammendrag
In a world where people are more connected, the barriers between deaf people and hearingpeople is more visible than ever. A neural sign language translation system would breakmany of these barriers. However, there are still many tasks to be solved before full automaticsign language translation is possible. Sign Language Translation is a difficult multimodalmachine translation problem with no clear one-to-one mapping to any spoken language.In this paper I give a review of sign language and its challenges regarding neural machinetranslation. I evaluate the state-of-the-art Sign Language Translation approach, and applya modified version of the Evolved Transformer to the existing Sign Language Transformer.I show that the Evolved Transformer encoder produces better results over the Transformerencoder with lower dimensions.