Papers by SILKNOW EU
Wissenschaftlich-Technische Jahrestagung der DGPF und Dreiländertagung der OVG, 2019
Der vorliegende Beitrag befasst sich mit der Klassifikation von Bildern mittels Deep Learning. Ex... more Der vorliegende Beitrag befasst sich mit der Klassifikation von Bildern mittels Deep Learning. Exemplarisch wird hier die bildbasierte Prädiktion der Entstehungs-epoche von Seidenstoffen und Fertigungsskizzen betrachtet. An diesem Beispiel wird die Problemstellung von nicht eindeutig definierten Klassen sowie uneindeutigen Trainingsbei-spielen aufgegriffen und ein Lösungsansatz vorgestellt. Dieser basiert auf einer erweiterten Verlustfunktion, mit welcher die uneindeutigen Trainingsbeispiele während des Trainings genutzt werden können. Hierzu werden während der Trainingsphase übergreifende Klassen definiert. Die durchgeführten Experimente zeigen, dass hiermit die Genauigkeit der Klassi-fikation ohne die übergreifenden Klassen verbessert werden kann. Da der exemplarische Datensatz relativ wenig Daten umfasst wird ein vortrainiertes Convolutional Neural Network verwendet und eine Augmentierung der Daten durchgeführt. Der Einfluss der Aug-mentierung sowie der entwickelten Verlustfunktion wird in einer Kreuzvalidierung evaluiert.
Heritage , 2019
Historical weaving techniques have evolved in time and space giving as result more or less fabric... more Historical weaving techniques have evolved in time and space giving as result more or less fabrics with different aesthetical characteristics. These techniques were transferred along the main silk production centers, thanks to the European Silk Road and creating a common European Frame on themes and techniques. These had made it complicated to determine whether a fabric corresponds to one century or another. Moreover, in order to understand their creation, it is necessary to determine the number of weaves and interlacements that each textile has, therefore, mathematical models can be extracted from these layers. In this sense, three dimensional (3D) virtual representations of the internal structure of textiles are of interest for a variety of purposes related to fashion, industry, education or other areas. The aim of this paper is to propose a mathematical modelling of historical weaving techniques by means of matrices in order to be easily mapped to a virtual 3D representation. The work focuses on historical silk textiles, ranging from the 15th to the 19th centuries. We also propose a computer vision-based strategy to extract relevant information from digital imagery, by considering different types of images (textiles, technical drawings and macro images). The work here presented has been carried out in the scope of the SILKNOW project, which has received funding from the European Union's Horizon 2020 research and innovation program under grant agreement No. 769504. The results shown in the paper are preliminary and will be improved in the scope of the project.
Symmetry, 2020
Symmetry is part of textile art in patterns and motifs that decorate fabrics, which are made by t... more Symmetry is part of textile art in patterns and motifs that decorate fabrics, which are made by the interlacement of warp and wefts. Moreover, the 3D representation of fabrics have already been studied by some authors; however, they have not specifically dealt with preserving historical weaving techniques. In this paper, we present the SILKNOW's Virtual Loom, a tool intended to document, preserve and reproduce silk historical weaving techniques from the 15th to the 19th centuries. We focus on the symmetry function and its contribution to art history, textile conservation, and modern design. We analyzed 2028 records from Garin 1820 datasets-a historical industry that still weaves with these techniques-and we reconstructed some historical designs that presented different types of defects. For those images (including fabrics and drawings) that had a symmetrical axis, we applied the symmetry functionality allowing to reconstruct missing parts. Thanks to these results, we were able to verify the usefulness of the Virtual Loom for conservation, analysis and new interpretative advantages, thanks to symmetry analysis applied to historical fabrics.
DATATEXTIL, 2020
La historia europea está tejida en seda. Aunque se suele asociar la Ruta de la Seda a sus orígene... more La historia europea está tejida en seda. Aunque se suele asociar la Ruta de la Seda a sus orígenes asiáticos, sus ramificaciones europeas fueron fundamentales para la construcción de la Europa actual gracias a los numerosos intercambios comerciales derivados de dichas rutas. La herencia de estas rutas ha dejado un legado incalculable: un ejemplo único de patrimonio donde la memoria, la identidad, la creatividad y el conocimiento confluyen en una única pieza. Es además un patrimonio vivo, multifacético en el que tradición y artesanía perviven. Sin embargo, a pesar de su enorme importancia en la construcción de la Europa contemporánea, este patrimonio no ha sido lo suficientemente valorado, esto aunado a su propia naturaleza física lo hace más frágil frente a otro tipo de elementos culturales como pinturas o esculturas. Más aún, cuando el patrimonio inmaterial relacionado con las piezas textiles está gravemente amenazado por el cierre de muchas industrias, la pérdida de memoria oral, la inminente desaparición de los conocedores de las técnicas tradicionales, etc. Esto resulta en que la fragilidad de sus materiales 1 y saberes vinculados con el diseño textil necesiten del uso de la tecnología para la conservación y difusión de este patrimonio. Proyecto de investigación SILKNOW, tejiendo el pasado hacia el futuro
PH: Boletín del Instituto Andaluz del Patrimonio Histórico, 2020
El patrimonio textil es uno de los varios "primos pobres" dentro de la gran familia del patrimoni... more El patrimonio textil es uno de los varios "primos pobres" dentro de la gran familia del patrimonio cultural. Algo especial-mente lamentable, pues pocas partes del registro histórico o cultural son tan transversales como los tejidos, tan ricas en información social y connotaciones culturales, técnicas y vitales. Tradicionalmente la seda ha tenido un lugar principal, como tejido de lujo, cuyos productos eran el máximo de las posibilidades materiales y culturales de muchas socieda-des. Esto hace también que los objetos de seda sean especialmente abundantes en muchas colecciones históricas; no solo por la indumentaria, sino como parte de objetos funcionales, simbólicos, litúrgicos, mobiliario, etc. Esas colecciones, a veces en grandes instituciones de rango nacional o global, a veces en pequeñas entidades locales, han sido documentadas de formas muy diversas a lo largo del tiempo. Tras décadas de informatización en los museos, no son pocos los que cuentan con algún tipo de repositorio digital de sus colecciones. A este respecto, la diversidad de situaciones es abrumadora: desde un sencillo listado en una hoja de cálculo a todo un sistema de gestión de coleccio-nes en la nube. En el caso del patrimonio textil, esa información ya es abundante en soporte digital, pero normalmente se queda dentro de los muros de las instituciones, al alcance solamente de profesionales y especialistas. La propia naturaleza minori-taria de este patrimonio, antes citada, es causa y consecuencia de esta situación. Fichas de catalogación antiguas, de autoría y autoridad dudosas, poco o nada normalizadas en su estructura y su terminología, según metodologías muy diversas, en idiomas diversos… Las instituciones museísticas son custodios de un valioso patrimonio textil; no obstante, es difícil hacerlo atractivo para los canales masivos, pues difícilmente entra en las categorías más habituales en historia del arte, como una autoría asociada a un nombre famoso, una ubicación muy reconocida, o una gran visibilidad pública. En ese sentido, una forma de sacar los tejidos de los expositores y cajones donde son-afortunadamente-preservados es hacerlos máximamente disponibles a través de medios digitales; no ya solo las piezas excepcionales, sino los conjuntos, las series tipológicas, que en estos casos permiten hilar discursos mucho más ricos y significativos. La accesibilidad también ha de ser para el público no especialista o con dificultades de visión o de comprensión intelectual. Desde SILKNOW (Silk heritage in the Knowledge Society: from punched cards to big data, deep learning and visual/ tangible simulations) nos hemos propuesto, entre otras cosas, mejorar el acceso a ese patrimonio y la información digital asociada, mediante diversas herramientas. Como todo proyecto reciente del programa Horizonte2020, seguimos una política de acceso abierto a los datos que recopilemos. Provienen de distintas fuentes: fundamentalmente, bases de datos internas facilitadas por museos y colecciones, con los que llegamos a acuerdos; y datos ya en abierto, co-sechados mediante API o a través de las sedes web. El resultado, todavía en construcción, será un repositorio donde aunar información de muchas fuentes y hacerlas disponibles a los usuarios mediante herramientas mejoradas: interfaz e interrogación multilingüe, gracias a un tesauro especializado, organizando la información mediante un modelo de datos específico para este tipo de patrimonio, basado en CIDOC-CRM. En esta tarea, apoyamos y proponemos enfo-ques de acceso abierto a la cultura, concretamente a la catalogación de las colecciones, tal como hacen Europeana o Wikimedia Commons. Para ello, y dado el tema de este debate, es necesaria una mayor difusión de la cultura de acceso abierto dentro del sector patrimonial y museístico. Las colecciones, los datos en sus catálogos, son la materia prima, no solamente para un proyecto como el nuestro, sino para cualquier desarrollo digital serio en las instituciones del patrimonio. Aunque al-gunos grandes museos están marcando el camino a seguir (en la estela del Rijksmuseum o el Metropolitan), haciendo libremente disponibles grandes cantidades de datos sobre sus colecciones, a otros les está costando seguir el mismo camino. Nuestro Museo del Prado, puntero en algunos aspectos tecnológicos (con su Archivo Digital o su Modelo Se
European Journal of Science and Theology, 2020
Clothes and textiles make up a very relevant part or religious cultural heritage. This paper pres... more Clothes and textiles make up a very relevant part or religious cultural heritage. This paper presents a selection of liturgical textiles from the 18 th and 19 th centuries. They were created by Garín, a Spanish factory still active today. The designs and weaving techniques employed in them have provided the starting point for a research project, SILKNOW, in operation between 2018 and 2021. It aims to apply cutting-edge computing technologies to textile heritage, including the religious and liturgical, and thus establish new historical and artistic connections.
cience and Digital Technology for Cultural Heritage-Interdisciplinary Approach to Diagnosis, Vulnerability, Risk Assessment and Graphic Information Models: Proceedings of the 4th International Congress Science and Technology for the Conservation of Cultural Heritage (TechnoHeritage 2019), 2019
The cultural heritage domain in general and silk textiles in particular, are characterized by lar... more The cultural heritage domain in general and silk textiles in particular, are characterized by large, rich and heterogenous data sets. Silk heritage vocabulary comes from multiple sources that have been mixed up across time and space. This has led to the use of different terminology in specialized organizations in order to describe their artifacts. This makes data interoperability between independent catalogues very difficult. Moreover, the interaction level of existing resources is low, most complex queries are not possible and results are poorly shown. In this regard, a recent EU-funded research project titled SILKNOW is building a multilingual thesaurus related to silk textiles. It is being carried out by experts in textile terminology and art historians, and computationally implemented by experts in text mining, multi-/cross-linguality and semantic extraction from text. This paper presents the rationale behind the realization of this thesaurus.
Esta contribución se enmarca dentro del proyecto H2020 Silknow que coordina la Universitat de Val... more Esta contribución se enmarca dentro del proyecto H2020 Silknow que coordina la Universitat de València a través de un equipo multidisplicinar conformado por investigadores de las TIC y del ámbito de las humanidades y de las ciencias sociales. Hoy el concepto de patrimonio cultural está más que nunca ligado al presente, como punto de partida que nos vincula con nuestro pasado cultural a través del acto consciente de preservar y legar a las generaciones futuras, convirtiendo a la sociedad en custodia. Este acto emerge con la ambición de permanecer en un mundo cada vez más cambiante y mutable, pero en medio de esta corriente surge cada vez con más presencia la apreciación del patrimonio no solo a través de su poder comunicativo que nos liga con nuestro pasado, sino también con su potencial como propiciador de economías innovadoras, a través del desarrollo sostenible y el impulso de las industrias creativas. Centrado en el patrimonio de la seda, se analiza este pasado desde el presente y como motor de desarrollo futuro.
Heritage, 2019
Historical weaving techniques have evolved in time and space giving as result more or less fabric... more Historical weaving techniques have evolved in time and space giving as result more or less fabrics with different aesthetical characteristics. These techniques were transferred along the main silk production centers, thanks to the European Silk Road and creating a common European Frame on themes and techniques. These had made it complicated to determine whether a fabric corresponds to one century or another. Moreover, in order to understand their creation, it is necessary to determine the number of weaves and interlacements that each textile has, therefore, mathematical models can be extracted from these layers. In this sense, three dimensional (3D) virtual representations of the internal structure of textiles are of interest for a variety of purposes related to fashion, industry, education or other areas. The aim of this paper is to propose a mathematical modelling of historical weaving techniques by means of matrices in order to be easily mapped to a virtual 3D representation. The work focuses on historical silk textiles, ranging from the 15th to the 19th centuries. We also propose a computer vision-based strategy to extract relevant information from digital imagery, by considering different types of images (textiles, technical drawings and macro images). The work here presented has been carried out in the scope of the SILKNOW project, which has received funding from the European Union's Horizon 2020 research and innovation program under grant agreement No. 769504. The results shown in the paper are preliminary and will be improved in the scope of the project.
Springer, 2019
Three dimensional (3D) virtual representations of the internal structure of textiles are of inter... more Three dimensional (3D) virtual representations of the internal structure of textiles are of interest for a variety of purposes related to fashion, industry, education or other areas. The modeling of ancient weaving techniques is relevant to understand and preserve our heritage, both tangible and intangible. However, ancient techniques cannot be reproduced with standard approaches, which usually are aligned with the characteristics of modern, mechanical looms. The aim of this paper is to propose a mathematical modelling of ancient weaving techniques by means of matrices in order to be easily mapped to a virtual 3D representation. The work focuses on ancient silk textiles, ranging from the 15th to the 19th centuries. We also propose a computer vision-based strategy to extract relevant information from digital imagery, by considering different types of images (textiles, technical drawings and macro images). The work here presented has been carried out in the scope of the SILKNOW project, which has received funding from the European Union's Horizon 2020 research and innovation program under grant agreement No. 769504.
Springer, 2019
Interactive visualization of spatio-temporal data is a very active area that has experienced rema... more Interactive visualization of spatio-temporal data is a very active area that has experienced remarkable advances in the last decade. This is due to the emergence of fields of research such as big data and advances in hardware that allow better analysis of information. This article describes the methodology followed and the design of an open source tool, which in addition to interactively visualizing spatio-temporal data that are represented in an ontology, allows the definition of what to visualize and how to do it. The tool allows selecting, filtering and visualizing in a graphical way the entities of the ontology with spatiotemporal data, as well as the instances related to them. The graphical elements used to display the information are specified on the same ontology, extending the VISO graphic ontology, used for mapping concepts to graphic objects with RDFS/OWL Visualization Language (RVL). This extension contemplates the data visualization on rich real-time 3D environments, allowing different modes of visualization according to the level of detail of the scene, while also emphasizing the treatment of spatio-temporal data, very often used in cultural heritage models. This visualization tool involves simple visualization scenarios and high interaction environments that allow complex comparative analysis. It combines traditional solutions, like hypercube or time-animations with innovative data selection methods.
Quaderns de Filologia, 2018
Durante el siglo xviii, Valencia se convirtió en el principal centro manufac-turero español de te... more Durante el siglo xviii, Valencia se convirtió en el principal centro manufac-turero español de tejidos de seda. Con el fin de proporcionar modelos originales para los mismos se creó en el seno de la Academia de Bellas Artes de San Carlos la "Escuela de Flores, Ornatos y otros diseños adecuados para Tejidos" en 1784. El presente artículo atiende a sus precedentes, sus principales protagonistas y analiza su funcionamiento, además de considerar la relación que mantuvo a lo largo de su trayectoria con la indus-tria de la seda. Palabras clave: seda; pintura de flores; Academia de Bellas Artes; Valencia; diseño textil. Abstract: During the 18th century, Valencia turned in the main Spanish silk textile manufacturing center. In order to provide with original models for them, the "School of Flowers, Ornaments and other designs suitable for Tissues" was created within the Academy of Fine Arts of San Carlos in 1784. This article is about the precedents, the main protagonists and analyzes its ways of functioning, in addition to considering the relationship that she maintained throughout her history with the silk industry.
Multimodal Technologies Interact, 2018
Silk was a major factor for progress in Europe, mostly along the Western Silk Road’s network of p... more Silk was a major factor for progress in Europe, mostly along the Western Silk Road’s network of production and market centers. The silk trade also allowed for the exchange of ideas and innovations, having impacts at economic, technical, functional, cultural and symbolic levels. However, silk has today become a seriously endangered heritage. Although many European specialized museums are devoted to its preservation, they usually lack the size and resources to take advantage of state-of-the-art digital technologies. The aim of this paper is twofold; firstly, we introduce SILKNOW, an interdisciplinary project that has been recently funded by the H2020 Programme of the European Union in order to preserve and promote the heritage of silk textiles; secondly, we introduce a set of interactive tools related to the project.
Conference Presentations by SILKNOW EU
ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., IV-2/W6, 2019
This paper presents a method for the classification of images of silk fabrics with the aim to pre... more This paper presents a method for the classification of images of silk fabrics with the aim to predict properties such as the place and time of origin and the production technique. The proposed method was developed in the context of the EU project SILKNOW (http://silknow.eu/). In the context of classification, we address the problem of limited as well as not fully labelled data and investigate the connection between the distinct variables. A pre-trained Convolutional Neural Network (CNN) is used for the feature extraction and a classification network realizing Multi-task learning (MTL) is trained based on these features. The training procedure is adapted to enable the consideration of images that do not have a label for all tasks. Additionally, MTL with fully labeled training data is investigated for the classification of silk fabrics. The impact of both MTL approaches is compared to single-task learning based on two different class structures. We achieve overall accuracies of 92-95% and average F1-scores of 88-90% in our best experiments.
We present results of collaborative work bringing together semantic technologies, machine learnin... more We present results of collaborative work bringing together semantic technologies, machine learning and cultural heritage to enable advanced search and visualization of textual descriptions of museum artifacts related to silk fabrics. Proposed is a multilingual txt analysis approach where the developed domain-specific multilingual thesaurus and domain-specific ontology are utilized in data representation and analysis. In addition, a general multilingual semantic annotation tool Wikifier is applied on thesaurus definitions and descriptions of silk-related museum artefacts. The validation on real-world data of several museums confirms suitability of the developed thesaurus and the ontology.
Uploads
Papers by SILKNOW EU
Conference Presentations by SILKNOW EU