Reconocimiento continuo de la Lengua de Señas Mexicana

Continuous recognition of Mexican Sign Language

Ricardo Fernando Morfín Chávez (2023, [Tesis de maestría])

La Lengua de Señas Mexicana (LSM) es la lengua utilizada por la comunidad Sorda en México, y, a menudo, subestimada y pasada por alto por la comunidad oyente, lo que resulta en la exclusión sistemática de las personas Sordas en diversos aspectos de la vida. Sin embargo, la tecnología puede desempeñar un papel fundamental en acercar a la comunidad Sorda con la comunidad oyente, promoviendo una mayor inclusión y comprensión entre ambas. El objetivo principal de este trabajo es diseñar, implementar y evaluar un sistema de reconocimiento continuo de señas estáticas en LSM mediante, visión por computadora y técnicas de aprendizaje máquina. Se establecieron objetivos específicos, que incluyen la generación de un conjunto de datos de señas estáticas, pertenecientes al alfabeto manual de la LSM, el diseño de un modelo de reconocimiento, y la evaluación del sistema, tanto en la modalidad aislada como en la continua. La metodología involucra dos evaluaciones distintas. La primera se enfoca en el reconocimiento de señas estáticas en el dominio aislado, para ello se capturaron datos de 20 participantes realizando movimientos de la mano en múltiples ángulos. Se evaluaron diversas técnicas de aprendizaje automático, destacando que el enfoque basado en Máquinas de Soporte Vectorial (SVM) obtuvo los mejores resultados (F1-Score promedio del 0.91). La segunda evaluación se concentra en el reconocimiento continuo de señas estáticas, con datos recopilados de seis participantes con diferentes niveles de competencia en LSM, logrando un rendimiento sólido con errores cercanos al 7 %. Además, se evaluó la viabilidad del sistema en aplicaciones de tiempo real, demostrando un excelente desempeño (velocidad promedio de procesamiento de 45 cuadros por segundo). A pesar de los logros alcanzados, es importante reconocer que este proyecto se centró en el reconocimiento continuo de señas estáticas en LSM. Queda pendiente, como un desafío interesante, la exploración del reconocimiento continuo de señas dinámicas en LSM para futuras investigaciones. Se considera esencial explorar enfoques orientados a la escalabilidad y aplicaciones en tiempo real en investigaciones posteriores.

This study focuses on the continuous recognition of static signs in Mexican Sign Language (Lengua de Señas Mexicana (LSM)), the language used by the Deaf community in Mexico. Despite its significance, LSM is often underestimated and overlooked, leading to the systematic exclusion of Deaf individuals in various aspects of life. The primary objective of this work is to design, implement, and evaluate a continuous static sign recognition system in LSM using computer vision and machine learning techniques. Specific goals were established, including the creation of a dataset of static signs belonging to the manual alphabet of LSM, the design of a recognition model, and the evaluation of the system in both isolated and continuous modes. The methodology involves two distinct evaluations. The first one focuses on the recognition of static signs in the isolated domain, for which data from 20 participants performing hand movements at various angles were collected. Various machine learning techniques were evaluated, with the Máquinas de Soporte Vectorial (SVM)-based approach achieving the best results (average F1-Score of 0.91). The second evaluation centers on the continuous recognition of static signs, using data collected from six participants with varying levels of competence in LSM, achieving robust performance with errors close to 7 %. Furthermore, the feasibility of the system in real-time applications was assessed, demonstrating excellent performance (average processing speed of 45 frames per second). Despite the achievements, it is important to recognize that this project focused on continuous recognition of static signs in LSM. It remains an interesting challenge to explore the continuous recognition of dynamic signs in LSM for future research. It is considered essential to explore scalability-oriented approaches and real-time applications in subsequent investigations.

Lengua de Señas Mexicana (LSM), visión por computadora, aprendizaje automático, alfabeto manual de la LSM, reconocimiento automático de señas estáticas, reconocimiento aislado de señas, reconocimiento continuo de señas, aplicacion Mexican Sign Language (LSM), computer vision, machine learning, LSM manual alpahbet, automatic recognition of static signs, isolated sign recognition, continuous sign recognition, real-time aplications INGENIERÍA Y TECNOLOGÍA CIENCIAS TECNOLÓGICAS TECNOLOGÍA DE LOS ORDENADORES ENSEÑANZA CON AYUDA DE ORDENADOR ENSEÑANZA CON AYUDA DE ORDENADOR

Offshore wind energy climate projection using UPSCALE climate data under the RCP8.5 emission scenario

MARKUS SEBASTIAN GROSS (2016, [Artículo])

In previous work, the authors demonstrated how data from climate simulations can be utilized to estimate regional wind power densities. In particular, it was shown that the quality of wind power densities, estimated from the UPSCALE global dataset in offshore regions of Mexico, compared well with regional high resolution studies. Additionally, a link between surface temperature and moist air density in the estimates was presented. UPSCALE is an acronym for UK on PRACE (the Partnership for Advanced Computing in Europe)-weather-resolving Simulations of Climate for globAL Environmental risk. The UPSCALE experiment was performed in 2012 by NCAS (National Centre for Atmospheric Science)- Climate, at the University of Reading and the UK Met Office Hadley Centre. The study included a 25.6-year, five-member ensemble simulation of the HadGEM3 global atmosphere, at 25km resolution for present climate conditions. The initial conditions for the ensemble runs were taken from consecutive days of a test configuration. In the present paper, the emphasis is placed on the single climate run for a potential future climate scenario in the UPSCALE experiment dataset, using the Representation Concentrations Pathways (RCP) 8.5 climate change scenario. Firstly, some tests were performed to ensure that the results using only one instantiation of the current climate dataset are as robust as possible within the constraints of the available data. In order to achieve this, an artificial time series over a longer sampling period was created. Then, it was shown that these longer time series provided almost the same results than the short ones, thus leading to the argument that the short time series is sufficient to capture the climate. Finally, with the confidence that one instantiation is sufficient, the future climate dataset was analysed to provide, for the first time, a projection of future changes in wind power resources using the UPSCALE dataset. It is hoped that this, in turn, will provide some guidance for wind power developers and policy makers to prepare and adapt for climate change impacts on wind energy production. Although offshore locations around Mexico were used as a case study, the dataset is global and hence the methodology presented can be readily applied at any desired location. © Copyright 2016 Gross, Magar. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reprod

atmosphere, climate change, Europe, Mexico, sampling, time series analysis, university, weather, wind power, climate, risk, theoretical model, wind, Climate, Models, Theoretical, Risk, Wind CIENCIAS FÍSICO MATEMÁTICAS Y CIENCIAS DE LA TIERRA CIENCIAS DE LA TIERRA Y DEL ESPACIO OCEANOGRAFÍA OCEANOGRAFÍA