Expert Systems publica un nuevo artículo del CEIEC: «ARQGAN: an evaluation of generative adversarial network approaches for automatic virtual inpainting restoration of Greek temples» // Expert Systems publishes a new CEIEC article: «ARQGAN: an evaluation of generative adversarial network approaches for automatic virtual inpainting restoration of Greek temples»

Los investigadores del CEIEC Alberto Nogales y Álvaro García Tejedor, junto con Emilio Delgado (Arquitectura, Universidad Francisco de Vitoria) y Ángel Melchor (alumni, Universidad Francisco de Vitoria) acaban de publicar el paper titulado «ARQGAN: an evaluation of generative adversarial network approaches for automatic virtual inpainting restoration of Greek temples» en la revista Expert systems with applicacions (Q1 en inteligencia artificial). En dicho trabajo se propone una metodología para la reconstrucción de templos griegos en ruinas representados en imágenes obtenidas de modelos 3D. Para ello se hace uso de modelos GAN que encontrarán los elementos arquitectónicos faltantes en la imagen, añadiéndolos automáticamente.

CEIEC researchers Alberto Nogales and Álvaro Garcia Tejedor, alongside Emilio Delgado (Architecture, Universidad Francisco de Vitoria) and Ángel Melchor (alumni, Universidad Francisco de Vitoria) have just published a paper entitled «ARQGAN: an evaluation of generative adversarial network approaches for automatic virtual inpainting restoration of Greek temples» in Expert Systems with Applications journal (Q1 in artificial intelligence). This work proposes a methodology to automatically rebuild remains of greek temples represented by images obtained from 3D models. To perform that, GAN models have been used which are able to find the missing architectonical elements of the image, adding them automatically.

Available at: https://www.sciencedirect.com/science/article/abs/pii/S0957417421005339