Machine Learning for Cultural Heritage: A Survey

Fiorucci, Marco; Khoroshiltseva, Marina; Pontil, Massimiliano; Traviglia, Arianna; Bue, Alessio Del; James, Stuart

doi:10.1016/j.patrec.2020.02.017

Cited by 193 publications

(117 citation statements)

References 27 publications

Supporting

Mentioning

117

Contrasting

Order By: Relevance

“…Several discarded features, products of recent anthropogenic activities, were also stored in separate libraries as false-positive traces. Due to their similarity to the ancient structures, they could contribute to training computer-assisted detection systems in future machine-learning projects [124][125][126].…”

Section: Sourcementioning

confidence: 99%

Following the Roman Army between the Southern Foothills of the Cantabrian Mountains and the Northern Plains of Castile and León (North of Spain): Archaeological Applications of Remote Sensing and Geospatial Tools

Blanco¹,

Sánchez

García

et al. 2020

Geosciences

View full text Add to dashboard Cite

Sixty-six new archaeological sites have been discovered thanks to the combined use of different remote sensing techniques and open access geospatial datasets (mainly aerial photography, satellite imagery, and airborne LiDAR). These sites enhance the footprint of the Roman military presence in the northern fringe of the River Duero basin (León, Palencia, Burgos and Cantabria provinces, Spain). This paper provides a detailed morphological description of 66 Roman military camps in northwestern Iberia that date to the late Republic or early Imperial eras. We discuss the different spatial datasets and GIS tools used for different geographic contexts of varied terrain and vegetation. Finally, it stresses out the relevance of these novel data to delve into the rationale behind the Roman army movements between the northern Duero valley and the southern foothills of the Cantabrian Mountains. We conclude that methodological approaches stimulated by open-access geospatial datasets and enriched by geoscientific techniques are fundamental to understand the expansion of the Roman state in northwestern Iberia during the 1st c. BC properly. This renewed context set up a challenging scenario to overcome traditional archaeological perspectives still influenced by the cultural-historical paradigm and the pre-eminence of classical written sources.

show abstract

Section: Sourcementioning

confidence: 99%

Following the Roman Army between the Southern Foothills of the Cantabrian Mountains and the Northern Plains of Castile and León (North of Spain): Archaeological Applications of Remote Sensing and Geospatial Tools

Blanco¹,

Sánchez

García

et al. 2020

Geosciences

View full text Add to dashboard Cite

show abstract

“…The results in this study have demonstrated the importance of a MAS approach to large-format paintings and the value of our two novel software tools for art conservation studies. In further research, we believe that further software tools could be developed to reveal the artist’s intentions and creative process, for example, using machine learning techniques similar to deep neural networks 60 .…”

Section: Discussionmentioning

confidence: 99%

Revealing time’s secrets at the National Theatre of Costa Rica via innovative software for cultural heritage research

Barrantes-Madrigal

Zúñiga-Salas

Arce-Tucker

et al. 2021

Sci Rep

View full text Add to dashboard Cite

Establishing affordable, efficient, accessible, innovative, and multidisciplinary methodologies to the diagnosis of the conservation state of an artwork is key to carry out appropriate strategies of conservation and consequently to the creation of modern public policies on cultural heritage. Limited access to large-format paintings is a challenge to restoration scientists seeking to obtain information quickly, in a non-destructive and non-invasive manner, and identify regions of interest. Therefore, we put forward two unique software tools based on multispectral imaging techniques, with the long-term aim to assess the artist’s intentions, creative process, and colour palette. This development paves the way for a comprehensive and multidisciplinary understanding of the mysteries encompassed in each pictorial layer, through the study of their physical and chemical characteristics. We conducted the first ever study on Musas I and Musas II, two large-format paintings by Italian artist Carlo Ferrario, located in the National Theatre of Costa Rica. In this study, we used our novel imaging techniques to choose regions of interest in order to study sample layers; while also assessing the works’ state of conservation and possible biodeterioration. We explored the applications of our two versatile software tools, RegionOfInterest and CrystalDistribution, and confirmed paint stratigraphies by means of microscopy and spectroscopy analyses (OM, SEM-EDX, Fluorescent microscopy, FTIR-ATR and micro-Raman). In a pilot study, we identified the artist’s main colour palette: zinc white, lead white, chrome yellow, lead read, viridian, along with artificial vermilion and ultramarine pigments. We were able to identify artificial vermilion and ultramarine and distinguish them from the natural pigments using CrystalDistribution to map the average size and diameter of the pigment crystals within the paint layers. This study demonstrated that software-based multidisciplinary imaging techniques are novel in establishing preventive and non-invasive methods for historical painting conservation studies, in addition, this study provides tools with great potential to be used in the future in applications such as virtual restoration.

show abstract

“…Fiorucci et al [16] argues that due to the 'black box' nature of the modern DL classifiers, they are being slowly implemented to cultural heritage applications but pioneers in this field aim to make DL methods a standard tool in this domain [17][18][19][20][21]. LiDAR sensing produces large amounts of data that are processed with the aid of a wide range of computer vision methods.…”

Section: Overview Of Deep Learning Methods For Semantic Segmentationmentioning

confidence: 99%

Semantic Segmentation of Airborne LiDAR Data in Maya Archaeology

et al. 2020

View full text Add to dashboard Cite

Airborne LiDAR produced large amounts of data for archaeological research over the past decade. Labeling this type of archaeological data is a tedious process. We used a data set from Pacunam LiDAR Initiative survey of lowland Maya region in Guatemala. The data set contains ancient Maya structures that were manually labeled, and ground verified to a large extent. We have built and compared two deep learning-based models, U-Net and Mask R-CNN, for semantic segmentation. The segmentation models were used in two tasks: identification of areas of ancient construction activity, and identification of the remnants of ancient Maya buildings. The U-Net based model performed better in both tasks and was capable of correctly identifying 60–66% of all objects, and 74–81% of medium sized objects. The quality of the resulting prediction was evaluated using a variety of quantifiers. Furthermore, we discuss the problems of re-purposing the archaeological style labeling for production of valid machine learning training sets. Ultimately, we outline the value of these models for archaeological research and present the road map to produce a useful decision support system for recognition of ancient objects in LiDAR data.

show abstract

Machine Learning for Cultural Heritage: A Survey

Cited by 193 publications

References 27 publications

Following the Roman Army between the Southern Foothills of the Cantabrian Mountains and the Northern Plains of Castile and León (North of Spain): Archaeological Applications of Remote Sensing and Geospatial Tools

Following the Roman Army between the Southern Foothills of the Cantabrian Mountains and the Northern Plains of Castile and León (North of Spain): Archaeological Applications of Remote Sensing and Geospatial Tools

Revealing time’s secrets at the National Theatre of Costa Rica via innovative software for cultural heritage research

Semantic Segmentation of Airborne LiDAR Data in Maya Archaeology

Contact Info

Product

Resources

About