Pose-based deep learning approaches for recognizing isolated signs in Greek sign language
View/
Keywords
Greek Sign Language (GSL) ; Sign language recognition ; Isolated gloss recognition ; Hand skeletal landmarks ; Sequence classification ; Deep learning ; Long short-term memory ; Graph convolutional networksAbstract
This thesis explores the task of isolated sign recognition in Greek Sign Language (GSL) using deep learning. GSL, like many sign languages, lacks large-scale annotated datasets, making automatic recognition a challenging problem. To address this, we use the publicly available GSL RGB+D dataset, which contains annotated video recordings captured with an Intel RealSense depth camera. We implement and evaluate three distinct neural architectures: a Convolutional Neural Network (CNN), a Long Short-Term Memory (LSTM) network, and a Graph Convolutional Network (GCN). Each model is designed to handle different characteristics of sign language data, visual, temporal, and spatial. Our experiments, conducted on the isolated gloss subset of the dataset, show that the LSTM model achieves the highest overall accuracy, while the CNN and GCN models demonstrate strength in specific categories. These findings underline the importance of temporal and structural information in sign recognition. This work contributes a comparative study of recognition models tailored to Greek Sign Language and highlights their potential in low-resource language contexts.
Postgraduate Studies Programme
Τεχνητή Νοημοσύνη - Artificial IntelligenceDepartment
Σχολή Τεχνολογιών Πληροφορικής και Επικοινωνιών. Τμήμα Ψηφιακών ΣυστημάτωνCorporate Department
National Center of Scientific Research "Demokritos"Number of pages
73Language
EnglishCollections
Except where otherwise noted, this item's license is described as
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα