The week's pick
Random Articles
Reseach Article
Signify: A Real-Time Sign to Text and Text to Sign Mobile Application for Dynamic Filipino Sign Language Translation using Transformer Architecture Deep Learning Model
| International Journal of Computer Applications |
| Foundation of Computer Science (FCS), NY, USA |
| Volume 187 - Number 87 |
| Year of Publication: 2026 |
| Authors: Aliyah Ayco, Kaye Anne Mirador, Glaiza Mei Natividad, Noah Andrea Pagba, James Esquivel |
10.5120/ijca2026926514
|
Aliyah Ayco, Kaye Anne Mirador, Glaiza Mei Natividad, Noah Andrea Pagba, James Esquivel . Signify: A Real-Time Sign to Text and Text to Sign Mobile Application for Dynamic Filipino Sign Language Translation using Transformer Architecture Deep Learning Model. International Journal of Computer Applications. 187, 87 ( Mar 2026), 38-44. DOI=10.5120/ijca2026926514
Abstract
This study presents Signify, a real-time, bidirectional mobile application for dynamic Filipino Sign Language (FSL) translation designed to bridge communication gaps between the Deaf and Hard of Hearing (DHH) community and hearing individuals. Utilizing Long Short-Term Memory (LSTM) and Transformer architectures, the system enables Sign-to-Text (S2T) and Text-to-Sign (T2S) functionalities. To improve model robustness, the researchers expanded the FSL-105 dataset by adding a "Directions" category and recording 80 additional videos per gesture, resulting in a total of 11,530 videos. For S2T recognition, hand landmarks were extracted via MediaPipe. Comparative analysis revealed that the Transformer model significantly outperformed the LSTM baseline, achieving a test accuracy of 98.73%. This was further improved to 99.60% through data augmentation techniques including Gaussian noise injection and temporal jitter. The T2S module utilizes a direct mapping approach to retrieve pre-recorded FSL video segments validated by a certified interpreter for linguistic accuracy. Integrated into an Android application using TensorFlow Lite, the system supports real-time, offline inference. User usability testing yielded a Grand Overall Mean of 4.57 (Excellent), reflecting high satisfaction among signers and non-signers. This research advances inclusive communication in alignment with SDGs 4 and 10.
References
- Hand Talk. 2023. The Benefits of Sign Language for Children with Hearing Loss. Hand Talk - Learn ASL today.
- Tolentino, L. K., Serfa Juan, R., Thio-ac, A., Pamahoy, M. A., Forteza, J. R. and Garcia, X. J. 2019. Static Sign Language Recognition Using Deep Learning. International Journal of Machine Learning and Computing 9(6), 821-827.
- Sundar, B. and Bagyammal, T. 2022. American Sign Language recognition for alphabets using MediaPipe and LSTM. Procedia Computer Science 215, 642-651.
- Canlas et al. 2024. Real-time dynamic Filipino Sign Language recognition application. Angeles University Foundation.
- Chaudhary, L., Ananthanarayana, T., Hoq, E. and Nwogu, I. 2023. SignNet II: A transformer-based two-way sign language translation model. IEEE Transactions on Pattern Analysis and Machine Intelligence 45(11), 12896-12907.
- Tupal, J. 2023. FSL-105 Dataset. Mendeley Data repository.
- Evangelista, C. L. L., Geli, C. J. R., Castillo, M. M. V. and Macabagdal, C. B. G. 2023. Long Short-Term Memory-based Static and Dynamic Filipino Sign Language Recognition. IEEE.
- Nerlekar, A. 2021. Sign Language Recognition Using Smartphones. California State University.
- Caya, M. V. C., Madrid, G. K. R. and Villanueva, R. G. R. 2022. Recognition of Dynamic Filipino Sign Language using MediaPipe and Long Short-Term Memory. 13th ICCCNT.
- Kothadiya, D., Bhatt, C., Sapariya, K., Patel, K., Gil-González, A.-B. and Corchado, J. M. 2022. Deepsign: Sign Language Detection and Recognition Using Deep Learning. Electronics 11(11), 1780.
- Do Long, V. 2021. Mobile Application for Sign Language Recognition. Czech Technical University.
- De Coster, M., Van Herreweghe, M. and Dambre, J. 2020. Sign language recognition with transformer networks. LREC, 6018-6024.
- Abdul, W., Alsulaiman, M., Amin, S. U., Faisal, M., Muhammad, G., Albogamy, F. R. and Ghaleb, H. 2021. Intelligent real-time Arabic sign language classification using attention-based inception and BiLSTM. Computers & Electrical Engineering 95, 107395.
- Ananthanarayana, T. 2021. A Comprehensive Approach to Automated Sign Language Translation. Rochester Institute of Technology.
- Wei, S. and Lan, Y. 2023. A two-way translation system of Chinese sign language based on computer vision. arXiv preprint.
- Stoll, S., Camgoz, N. C., Hadfield, S. and Bowden, R. 2020. Text2Sign: Towards sign language production using neural machine translation and generative adversarial networks. International Journal of Computer Vision 128(6), 891–908.
- toll, S., Hadfield, S. and Bowden, R. 2020. Signsynth: Data-driven sign language video generation. ECCV, 353-370.
- Latkar, H., Wasker, S., Vashistha, A. and Kanse, A. 2024. Real-time conversion of sign language to text and speech, and vice-versa. TIJER 11(4).
- Faisal, M., Alsulaiman, M., Mekhtiche, M., Abdelkader, B. M., Algabri, M. and Alrayes, T. B. S. 2023. Enabling two-way communication of deaf using Saudi sign language. IEEE Access 11, 135423-135434.
- Kavana and Suma. 2022. Real-time sign language recognition.
- Manoj Kumar, D., Bavanraj, K., Thavananthan, S., Bastiansz, G. M. A. S., Harshanath, S. M. B. and Alosious, J. 2020. EasyTalk: A translator for Sri Lankan Sign Language using machine learning. ICAC.
- Saleem, M. I., Siddiqui, A., Noor, S., Luque-Nieto, M. A. and Otero, P. 2022. A novel machine learning based two-way communication system for deaf and mute. Applied Sciences 13(1), 453.
- Ramadhan, M. F., Samsuryadi, S. and Primanita, A. 2024. American sign language translation to display the text (subtitles) using a convolutional neural network. EMACS Journal 6(3).
- Osman et al. 2020. Hearing Assistive Technology: Sign Language Translation Application for Hearing-Impaired Communication.
- Kautsar et al. 2024. Transformer-based sequence models.
Index Terms
Keywords