AzSLD - Azerbaijani Sign Language Dataset

The Azerbaijani Sign Language Dataset (AzSLD) is a comprehensive, multi-modal dataset designed to facilitate the development and evaluation of machine learning models for the recognition and translation of Azerbaijani Sign Language (AzSL). This dataset is specifically tailored for use in sign language recognition systems and other applications requiring accurate interpretation of dactylology, word-level, and sentence-level signs in AzSL.

AzSLD is the first publicly available dataset focused on Azerbaijani Sign Language. It contributes to the global effort to improve accessibility for the deaf and hard-of-hearing community in Azerbaijan. The dataset aims to bridge the gap between technology and accessibility by providing high-quality data for researchers, developers, and practitioners working on sign language recognition systems.

Dataset Components: AzSLD is organized into three primary components:

• AzSLD_Sentences: Contains video sequences of complete sentences in AzSL. This component is designed to capture the fluidity and contextual nature of sign language, providing data for more complex language modeling tasks. It includes detailed annotations with timestamped glosses for each sentence, enabling precise analysis and model training.

• AzSLD_Words: Comprises a collection of videos representing commonly used words in AzSL. Each word is signed by multiple signers to account for variations in signing style, speed, and regional dialects. Videos are annotated with word labels in both Azerbaijani and English, facilitating use in bilingual and multilingual applications.

• AzSLD_Fingerspelling: Includes video sequences of individual handshapes representing the letters of the Azerbaijani alphabet. Each sign is captured from multiple angles to ensure comprehensive coverage of the dactylology in AzSL. This component is ideal for tasks involving letter recognition and the integration of fingerspelling into broader sign language recognition systems.

Key Features:

• Double-View Recordings: The dataset includes recordings from two camera angles to capture both frontal and side views of hand and body movements, ensuring that the subtle nuances of sign language are well-represented.

• Signers: The dataset features recordings from a diverse group of native AzSL signers, including variations in age, gender, and signing style. This diversity is crucial for training models that are robust to variations in signing.

• Annotations: Each video is annotated with detailed metadata, including the sign’s label (dactyl, word, or sentence), signer ID, and timestamped glosses for sentence-level signs.

• Data Format: The RGB videos are provided in high-definition (HD) resolution at 35 frames per second, with accompanying JSON files containing annotations and metadata. The dataset is organized into folders by category for easy navigation.

Applications:

• Sign Language Recognition: AzSLD is ideal for training and evaluating models designed to recognize and translate signs in Azerbaijani Sign Language.

• Machine Translation: The dataset can be used for developing machine translation systems that convert AzSL to written or spoken Azerbaijani.

• Linguistic Analysis: Researchers can utilize AzSLD to study the structure and usage patterns of AzSL, contributing to the documentation and preservation of the language.

• Educational Tools: The dataset can be employed in the development of educational software aimed at teaching AzSL to both deaf individuals and those interested in learning the language.

License: AzSLD is released under a Creative Commons Attribution 4.0 International license, allowing for academic and non-commercial use. Any commercial use requires explicit permission from the dataset creators.

Citation: When using AzSLD in your research, please cite the following paper:

Alishzade, N., Hasanov, J. (2024). AzSLD: Azerbaijani Sign Language Dataset for Dactyl, Word, and Sentence Translation with Baseline Software. [Journal Name], [Volume(Issue)], [Pages]. DOI: [DOI link].

Contact:
For questions, feedback, or contributions, please contact the project team at: slr.project.ada@gmail.com