We are excited to announce the release of Version 25.05.
Since 2023, our laboratory has been collecting a Japanese fingerspelling dataset.
To make this resource widely available to researchers, we are now releasing it under the name ub-MOJI.
Japanese fingerspelling is an essential means of communication for people who are deaf or hard of hearing, as well as for learners of Japanese Sign Language.
However, until now, there has been no large-scale and openly available dataset for research purposes.
This release marks the first initiative to systematically organize and share a Japanese fingerspelling dataset with the research community.
The initial release (v25.05) includes:
- Video and image data of fingerspelling performed by participants
- Character label information
- Basic metadata
We plan to expand the dataset with more participants and additional data types in future releases, making it suitable for a wider range of research applications.
We hope that ub-MOJI will support advances in research on fingerspelling recognition, education, and assistive technologies.