Deep Belief Networks for Multimodal, Images-based Non Contact Material Characterization
Journal Title: Digital Presentation and Preservation of Cultural and Scientific Heritage - Year 2017, Vol 7, Issue
Abstract
Our growing cognisance about the chemo-physical properties of elec-tromagnetic waves and of their interaction with various materials provides ex-panding range of possibilities for quantification and non contact material char-acterization even in difficult to computation, « non-standard » domains as this of the cultural heritage. This review - thematically part of the IFIDA project dedicated to digitization of archaeometric and conservation-restoration praxis - deals with the possibilities latest neuroinformatic methods offer for a more effi-cient and fast interrogation, understanding and classification of multi-modal spectral records of paintings on various supports. The workflow followed during their routine non destructive (ND) analysis by human experts is described in terms of specific technical characteristics, research objectives and concrete ap-plication for the artworks' characterization as a prototype to simulate artificially. References to techniques for high semantic level information extraction from low-level features are suggested. The strengths and limitations in application of Deep Belief Networks (DBNs) for compression, visualization and data recognition are also considered. Examples of most appropriate architectures and topological maps of Deep Learning Networks for interpretation of UV, IR, XR, β, γ and CT records performing simple characterization and classification tasks - from single layer perceptron to multi-layer deep belief neural networks for unlabelled data - are presented and explained.
Authors and Affiliations
Magdelena Stoyanova, Lilia Pavlova
Keywords
Related Articles
- EP ID EP326150
- DOI -
- Views 157
- Downloads 0
3D-Reconstruction of the Complex Stuccoes from Patrimony Buildings
The paper deals with 3D scanning techniques and instruments to survey the complex stuccoes from the architectural building Nanu-Muscel from Bucharest. EXAscan Portable 3D Laser Scanner has been used to exploit the advant...
Generation of Educational 3D Maze Games for Carpet Handicraft in Bulgaria
Serious video games applied for learning purposes play a significant and important role for the modern technology enhanced education. The paper presents an educational 3D maze video game dedicated to the development of c...
Accession of Unstructured File Collections
Unstructured file collections will keep archives increasingly busy, as the digitisation of the workplace as well as the private correspondence pro-ceeds. But the archiving of this kind of data requires a lot of time and...
Towards Building a Semantic Repository of Bioinformatics Resources
The paper presents a work in progress directed to the creation of a semantic digital repository of scholarly resources in the area of bioinformatics. A special attention has been paid to the design of a lightweight subje...
An E-Learning Environment – a Tool for Presentation of Knowledge about Bulgarian Cultural Heritage Sites
The article describes an interactive software environment for learning as an accessible tool to supplement the learning content. The environment, under development, is in line with the State Educational Standards (SES) o...