This is not the latest version of this item. The latest version can be found here.
Automatic SignWriting Recognition
dc.contributor.author | Sevilla, Antonio F. G. | |
dc.contributor.author | Díaz Esteban, Alberto | |
dc.contributor.author | Lahoz-Bengoechea, José María | |
dc.date.accessioned | 2023-06-15T06:21:25Z | |
dc.date.available | 2023-06-15T06:21:25Z | |
dc.date.issued | 2021-11 | |
dc.description.abstract | Sign languages are viso-gestual languages, using space and movement to convey meaning. To be able to transcribe them, SignWriting uses an iconic system of symbols meaningfully arranged in the page. This two-dimensional system, however, is very different to traditional writing systems, so its automatic processing poses a novel challenge for computational linguistics. We identify as first and fundamental step to overcome this challenge the extraction of a computational representation of the semantics represented by SignWriting transcriptions. We propose a data-based modelization of the problem, construed from real handwritten SignWriting instances. We then propose two solutions involving state of the art machine learning techniques combined with expert analysis. The first solution is direct application of an existing deep neural network. Our second proposal exploits the expert knowledge codified in the data annotation scheme that we present, in order to craft a system that improves on the straight-forward solution's accuracy by 30%. This improved system uses a number of different neural networks to divide the necessary processing, progressively constructing the prediction in an iterative pipeline that combines deep learning and domain knowledge in a mixed solution. | |
dc.description.abstract | Las lenguas de signos son lenguas viso-gestuales que utilizan el espacio y el movimiento para transmitir significado. Para transcribirlas, la SignoEscritura utiliza un sistema icónico de símbolos distribuidos de manera significativa por la página. Este sistema bidimensional es muy diferente a los sistemas de escritura tradicionales, lo que hace su tratamiento automático un desafío novedoso para la lingüística computacional. Identificamos como paso fundamental para superar este desafío la extracción de una representación computacional de la semántica representada por las transcripciones de SignoEscritura. Proponemos una modelización del problema basada en datos, obtenida a partir de ejemplos reales de SignoEscritura hecha a mano. Asimismo, proponemos dos posibles soluciones, utilizando técnicas del estado del arte de aprendizaje automático combinadas con análisis experto. La primera solución es la aplicación directa de una red neuronal profunda existente, mientras que nuestra segunda propuesta explota el conocimiento experto codificado en la anotación de los datos, anotación que a su vez presentamos, para crear un sistema que mejora la precisión respecto a la primera propuesta en un 30%. Este sistema mejorado utiliza una serie de distintas redes neuronales para dividir el procesamiento necesario, construyendo progresivamente la predicción en un proceso iterativo que combina el aprendizaje profundo con el conocimiento de dominio en una solución mixta. | |
dc.description.department | Depto. de Ingeniería de Software e Inteligencia Artificial (ISIA) | |
dc.description.department | Depto. de Lengua Española y Teoría de la Literatura | |
dc.description.refereed | FALSE | |
dc.description.sponsorship | Indra | |
dc.description.sponsorship | Fundación Universia | |
dc.description.status | unpub | |
dc.eprint.id | https://eprints.ucm.es/id/eprint/69235 | |
dc.identifier.uri | https://hdl.handle.net/20.500.14352/169 | |
dc.language.iso | eng | |
dc.relation.projectID | Visualizando la SignoEscritura (PR2014_19/01) | |
dc.rights.accessRights | open access | |
dc.subject.keyword | Sign Language | |
dc.subject.keyword | SignWriting | |
dc.subject.keyword | Deep Learning | |
dc.subject.keyword | Expert Knowledge | |
dc.subject.keyword | Neural Networks | |
dc.subject.keyword | Computer Vision | |
dc.subject.keyword | Lengua de Signos | |
dc.subject.keyword | SignoEscritura | |
dc.subject.keyword | Aprendizaje Profundo | |
dc.subject.keyword | Conocimiento Experto | |
dc.subject.keyword | Redes Neuronales | |
dc.subject.keyword | Visión Artificial | |
dc.subject.ucm | Inteligencia artificial (Informática) | |
dc.subject.ucm | Sistemas expertos | |
dc.subject.ucm | Informática (Filología) | |
dc.subject.unesco | 1203.04 Inteligencia Artificial | |
dc.title | Automatic SignWriting Recognition | |
dc.title.alternative | Reconocimiento Automático de SignoEscritura | |
dc.type | journal article | |
dcterms.references | Branchini, C., & Mantovan, L. (2020). A Grammar of Italian Sign Language (LIS). Fondazione Università Ca’ Foscari. doi:10.30687/978-88-6969-474-5. Eccarius, P., & Brentari, D. (2008). Handshape coding made easier: A theoretically based notation for phonological transcription. Sign Language & Linguistics, 11, 69–101. doi:10.1075/sll.11.1.11ecc. Granlund, G. H., & Knutsson, H. (1995). Signal processing for computer vision. Kluwer Academic Publishers. Hanke, T. (2004). HamNoSys – Representing Sign Language Data in Language Resources and Language Processing Contexts. In Proceedings of the Workshop on Representation and Processing of Sign Language, Workshop to the forth International Conference on Language Resources and Evaluation (LREC’04) (pp. 1–6). ISSN: 17913721. Herrero Blanco, Á. (2003). Escritura alfabética de la lengua de signos española: once lecciones. Publicaciones de la Universidad de Alicante. Herrero Blanco, Á. (2009). Gramática didáctica de la lengua de signos española (LSE). Ediciones SM. Hilzensauer, M., & Krammer, K. (2015). A multilingual dictionary for sign languages: “SpreadTheSign”. ICERI2015 Proceedings, (pp. 7826–7834). Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017). ImageNet classification with deep convolutional neural networks. Communications of the ACM 60, 84–90. doi:10.1145/3065386. Langer, J., Andres, J., Benešová, M., & Faltýnek, D. (2020). Quantitative lingustic analysis of Czech sign language. (1st ed.). Univerzita Palackého v Olomouci. doi:10.5507/pdf.20.24457277. Liddell, S. K., & Johnson, R. E. (1989). American Sign Language: The Phonological Base. Sign Language Studies, 1064, 195–277. doi:10.1353/sls. 1989.0027. O’Mahony, N., Campbell, S., Carvalho, A., Harapanahalli, S., Hernandez, G. V., Krpalkova, L., Riordan, D., & Walsh, J. (2020). Deep Learning vs. Traditional Computer Vision. In K. Arai, & S. Kapoor (Eds.), Advances in Computer Vision (pp. 128–144). Springer International Publishing. doi:10.1007/978-3-030-17795-9_10. Redmon, J., Divvala, S., Girshick, R., & Farhadi, A. (2016). You Only Look Once: Unified, Real-Time Object Detection. (pp. 779–788). URL: https://www.cv-foundation.org/openaccess/content_cvpr_2016/html/Redmon_You_Only_Look_CVPR_2016_paper.html. Redmon, J., & Farhadi, A. (2018). Yolov3: An incremental improvement. arXiv:1804.02767. Sambasivan, N., Kapania, S., Highfill, H., Akrong, D., Paritosh, P. K., & Aroyo, L. M. (2021). “Everyone wants to do the model work, not the data work”: Data Cascades in High-Stakes AI. In proceedings of the 2021 CHI Conference on Human Factors in Computing Systems (pp. 1–15). Sevilla, A. F. G., & Lahoz-Bengoechea, J. M. (2019). A different description of orientation in sign languages. Procesamiento del Lenguaje Natural, 62, (pp. 53–60). Smith, R. W. (2013). History of the Tesseract OCR engine: what worked and what didn’t. In Document Recognition and Retrieval XX (pp. 1–12). SPIE volume 8658. doi:10.1117/12.2010051. Stokoe, W. C. (1960). Sign Language Structure: An Outline of the Visual Communication Systems of the American Deaf. Studies in linguistics: Occasional papers, 8. Sutton, V. (2014). Lessons in sign writing: Textbook. (4th ed.). Center for Sutton Movement Writing. URL: https://www.signwriting.org/archive/docs2/sw0116-Lessons-SignWriting.pdf. Sutton, V., & Frost, A. (2008). SignWriting: sign languages are written languages!. Center for Sutton Movement Writing. | |
dspace.entity.type | Publication |
Download
Original bundle
1 - 1 of 1
Loading...
- Name:
- AutomaticSignWritingRecognition_SevillaDiazLahoz.pdf
- Size:
- 1.48 MB
- Format:
- Adobe Portable Document Format