A Comprehensive Study on Deep Learning-Based Methods for Sign Language Recognition

In this paper, a comparative experimental assessment of computer vision-based methods for sign language recognition is conducted. By implementing the most recent deep neural network in field, thorough evaluation on multiple publicly available datasets performed. The aim present study to provide insights recognition, focusing mapping non-segmented video streams glosses. For task, two new sequence training criteria, known from fields speech and scene text are introduced. Furthermore, plethora pretraining schemes thoroughly discussed. Finally, RGB+D dataset Greek created. To best our knowledge, first where sentence gloss level annotations provided capture.