Inteligencia Artificial en la educación artística: una perspectiva sobre su integración en la estética andina
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Rev. aportes de la comunicación
Abstract
Se propone la creación de una Red Generativa Adversarial (GAN), entrenada específica- mente con obras de estética andina para su revalorización y enseñanza, dado que la Inteligencia Artificial ha demostrado ser efectiva en estimular la motivación asociada en el ámbito educativo. Las tecnologías actualmente disponibles presentan sesgos en los resultados al generar imágenes que no están basadas en estilos ampliamente difundidos. Para ello, se estudian diferentes modelos técnicos, resaltando Sketch-And-Paint GAN de Alice Xue (2020) como una opción prometedora, además de opciones para la evaluación posterior de los resultados obtenidos. Finalmente, se subrayan algunos desafíos como la obtención de una base de datos adecuada y se resaltan las consideraciones éticas que implica el desarrollo de este tipo de tecnologías.
Currently, andean aesthetics experience a lack of recognition and appreciation in con- trast to their cultural and historical richness. Thus, the creation of a Generative Adversarial Ne- twork (GAN) specifically trained with works from this current is proposed for their revaluation and teaching. AI technologies have proven effective in stimulating motivation associated with their application in the educational field. However, the currently available technologies present biases in their results, generating images not based on widely disseminated styles. To address this, some technical models are studied, highlighting Alice Xue's Sketch-And-Paint GAN (2020) as a promising option, along with options for the subsequent evaluation of the obtained results. Finally, some challenges, such as obtaining an adequate database, are underscored, and the ethical considerations involved in developing such technologies are emphasized.
Currently, andean aesthetics experience a lack of recognition and appreciation in con- trast to their cultural and historical richness. Thus, the creation of a Generative Adversarial Ne- twork (GAN) specifically trained with works from this current is proposed for their revaluation and teaching. AI technologies have proven effective in stimulating motivation associated with their application in the educational field. However, the currently available technologies present biases in their results, generating images not based on widely disseminated styles. To address this, some technical models are studied, highlighting Alice Xue's Sketch-And-Paint GAN (2020) as a promising option, along with options for the subsequent evaluation of the obtained results. Finally, some challenges, such as obtaining an adequate database, are underscored, and the ethical considerations involved in developing such technologies are emphasized.
Description
No. 38