ALGSL89: An Algerian Sign Language Dataset

Öz

Automatic Sign Language Recognition (ASLR) is an area of active current research that aims to facilitate communication between deaf and hearing people. Recognizing sign language, particularly in the context of Algerian Sign Language (ALGSL), presents unique challenges that have yet to be comprehensively explored. So far, to the best of our knowledge, no study has considered the ALGSL Recognition. This is mainly due to the lack of available datasets. To overcome this challenge, we propose the ALGSL89 dataset, a pioneering effort in ALGSL research. The ALGSL89 dataset encompasses 4885 videos, capturing 89 distinct ALGSL signs, recorded by 10 subjects. This dataset serves as a foundational resource for advancing ASLR research specific to the Algerian signing community. In addition, we provide a comprehensive analysis of its characteristics, including statistical insights and detailed information on handshapes, positions, trajectories, and the dynamic aspects of sign movements. These details are crucial for researchers to gain a nuanced understanding of the dataset, ensuring its effective utilization in ASLR studies. In order to test the validity of our dataset, we provide the results obtained by applying a set of deep learning models. Finally, we present SignAtlas, an innovative ALGSL recognition system based on Autoencoder model.

Anahtar Kelimeler

• Algerian Sign Language
• Sign language recognition
• handshape recognition
• Autoencoder
• Deep learning

ALGSL89: Bir Cezayir İşaret Dili Veri Seti

Öz

Otomatik İşaret Dili Tanıma (ASLR), sağır ve işiten insanlar arasında iletişimi kolaylaştırmayı amaçlayan aktif bir araştırma alanıdır. Özellikle Cezayir İşaret Dili (ALGSL) bağlamında işaret dili tanıma, henüz kapsamlı bir şekilde incelenmemiş benzersiz zorluklar sunmaktadır. Bildiğimiz kadarıyla, şimdiye kadar ALGSL Tanıma üzerine bir çalışma yapılmamıştır. Bu durum, büyük ölçüde mevcut veri setlerinin eksikliğinden kaynaklanmaktadır. Bu zorluğun üstesinden gelmek için, ALGSL araştırmalarında öncü bir çaba olarak ALGSL89 veri setini öneriyoruz. ALGSL89 veri seti, 10 konu tarafından kaydedilen 89 farklı ALGSL işaretini kapsayan 4885 video içermektedir. Bu veri seti, Cezayir işaret dili topluluğuna özgü ASLR araştırmalarını ilerletmek için temel bir kaynak olarak hizmet etmektedir. Ek olarak, el şekilleri, pozisyonlar, yörüngeler ve işaret hareketlerinin dinamik yönleri dahil olmak üzere, karakteristiklerinin kapsamlı bir analizini sunuyoruz. Bu detaylar, araştırmacıların veri setini nüanslı bir şekilde anlamalarını ve ASLR çalışmalarında etkili bir şekilde kullanmalarını sağlamak için hayati öneme sahiptir. Veri setimizin geçerliliğini test etmek amacıyla, derin öğrenme modelleri uygulayarak elde ettiğimiz sonuçları sunuyoruz. Son olarak, Otoenkoder modeline dayanan yenilikçi bir ALGSL tanıma sistemi olan SignAtlas'ı sunuyoruz.

Anahtar Kelimeler

• Cezayir İşaret Dili
• İşaret Dili Tanıma
• El Şekli Tanıma
• Otoenkoder
• Derin Öğrenme.

Kaynakça

1. Abdelouafi, H. (2019). Teaching Sign Language to the Deaf Children in Adrar, Algeria.
2. Adeyanju, I., Bello, O., & Adegboye, M. (2021). Machine learning methods for sign language recognition: A critical review and analysis. Intelligent Systems with Applications, 12. https://doi.org/{https://doi.org/10.1016/j.iswa.2021.200056
3. AL-Qurishi, M., Khalid, T., & Souissi, R. (2021). Deep Learning for Sign Language Recognition: Current Techniques, Benchmarks, and Open Issues. IEEE Access, 9. https://doi.org/10.1109/ACCESS.2021.3110912
4. Alzubaidi, L., Zhang, J., Humaidi, A. J., Al-Dujaili, A. Q., Duan, Y., Al-Shamma, O., Santamaría, J., Fadhel, M. A., Al-Amidie, M., & Farhan, L. (2021). Review of deep learning: concepts, CNN architectures, challenges, applications, future directions. Journal of Big Data, 8.
5. Amal, D. (2016). Les points communs entre la Algerian Sign Language (LSA) - dialecte de Laghouat, Sud de l’Algérie - et la Langue des Signes Française (LSF). Licence thesis, « Acquisition et dysfonctionnement », Faculté ALLSHS d’Aix-en-Provence.
6. Chang, J., & Jin, S. (2017). An efficient implementation of 2D convolution in CNN. IEICE Electronics Express, 14, 4299-4308. https://doi.org/10.1587/elex.13.20161134
7. Helen, C., Brian, H., & Richard, B. (2011). Sign Language Recognition. Dans Visual Analysis of Humans: Looking at People (pp. 539 - 562). Springer.
8. Hicham, A. (2021, October). Deaf Education in Algeria: Is it a Sustainable Approach? Sociology Review, 5, 417–429.

9. Lanesman, S. (2016). Algerian Jewish Sign language: its emergence and survival. Masters thesis, University of Central Lancashire.
10. Lanesman, S., & Meir, I. (2012). Algerian Jewish Sign Language: A sociolinguistic sketch. Dans Sign Languages in Village Communities: Anthropological and Linguistic Insights (éd. 1, pp. 361–364). De Gruyter. http://www.jstor.org/stable/j.ctvbkjwzx.17
11. LeCun, Y., Bengio, Y., & Hinton, G. (2015, May). Deep Learning. Nature, 521, 436-44. https://doi.org/10.1038/nature14539
12. Lindemann, B., Müller, T., Vietz, H., Jazdi, N., & Weyrich, M. (2021). A survey on long short-term memory networks for time series prediction. Procedia CIRP, 99, 650-655. https://doi.org/https://doi.org/10.1016/j.procir.2021.03.088
13. Lugaresi, C., Tang, J., Nash, H., McClanahan, C., Uboweja, E., Hays, M., Zhang, F., Chang, C.-L., Yong, M. G., Lee, J., Chang, W.-T., Hua, W., Georg, M., & Grundmann, M. (2019). MediaPipe: A Framework for Building Perception Pipelines. MediaPipe: A Framework for Building Perception Pipelines.
14. Neidle, C., Thangali, A., & Sclaroff, S. (2012). Challenges in development of the American Sign Language Lexicon Video Dataset (ASLLVD) corpus. Proceedings of the LREC2012 5th Workshop on the Representation and Processing of Sign Languages: Interactions between Corpus and Lexicon. 143-150.
15. Nekkaa, F. (2015). Détection automatique de la main : Application à la reconnaissance de la langue des signes arabe. Master thesis, Systèmes Distribués et Méthodes Formelles (SDMF), Université Abdelhamid Mehri-Constantine 2.
16. Obi, Y., Claudio, K. S., Budiman, V. M., Achmad, S., & Kurniawan, A. (2022). Sign language recognition system for communicating to people with disabilities. Procedia Computer Science, 216, 13-20. https://doi.org/https://doi.org/10.1016/j.procs.2022.12.106
17. Rosete, F. R., Quiroga, F. M., Estrebou, C., Lanzarini, L., & Rosete, A. (2016). LSA64: An Argentinian Sign Language Dataset. ArXiv, abs/2310.17429, 794-803.
18. Ruben, R. (2005). Sign language: Its history and contribution to the understanding of the biological nature of language. Acta oto-laryngologica, 125, 464-467. https://doi.org/10.1080/00016480510026287
19. Sghier, M. M. (2007). Langage et surdité, descriptive de la langue des signes des sourds Oranais. Magistère thesis, Université d’Oran Es-Sénia.
20. Siriak, R., Skarga-Bandurova, I., & Boltov, Y. (2019). Deep Convolutional Network with Long Short-Term Memory Layers for Dynamic Gesture Recognition. 2019 10th IEEE International Conference on Intelligent Data Acquisition and Advanced Computing Systems: Technology and Applications (IDAACS), 1, pp. 158-162. https://doi.org/10.1109/IDAACS.2019.8924381
21. Srivastava, S., Gangwar, A., Mishra, R., & Singh, S. (2022). Sign Language Recognition System Using TensorFlow Object Detection API. Dans Communications in Computer and Information Science (pp. 634–646). Springer International Publishing. https://doi.org/10.1007/978-3-030-96040-7_48
22. Tomasz, K., Mariusz, O., Marian, W., & Dawid, W. (2015). Recognition of Hand Gestures Observed by Depth Cameras. International Journal of Advanced Robotic Systems, 12(4).
23. Vargas, L. P., Barba, L., Torres, C. O., & Mattos, L. (2011, January). Sign Language Recognition System using Neural Network for Digital Hardware Implementation. Journal of Physics: Conference Series, 274, 012051. https://doi.org/10.1088/1742-6596/274/1/012051
24. Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., & Polosukhin, I. (2017). Attention Is All You Need. CoRR.