An Automated Technique to Recognize and Extract Images from Scanned Archaeological Documents

Roullet, Cindy; Gauch, John M.; Vennarucci, Rhodora G.

doi:10.1109/icdarw.2019.00009

Cited by 3 publications

(1 citation statement)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…), as these terms are too imprecise and loosely defined for meaningful data-driven analysis. The project is in the process of developing a comprehensive analytic tool for art in Pompeii, the Pompei Pitture e Mosaici Explorer (PPMx) [Roullet et al 2019[Roullet et al , forthcoming 2021. This tool uses the digitized eleven volumes of Pompei: pitture e mosaici (PPM), the single most important resource for artwork in Pompeii, to train convolutional neural networks (CNNs) to recognize similarities among its plates, and it uses natural language processing (NLP) to parse its Italian captions.…”

Section: Introductionmentioning

confidence: 99%

Putting Space Syntax to the Test

Vennarucci

2021

SDH

Self Cite

View full text Add to dashboard Cite

While space syntax analysis has been widely applied to archaeological sites (including Pompeii), it is fundamentally limited by its isolation within the social sciences and its omission of decoration from the analysis of human cognition and movement within structures. At the same time, phenomenology in archaeology has typically arisen from the physical experiences of a limited number of professional archaeologists in a landscape, with little interest in digital embodiment in virtual spaces. The Virtual Pompeii Project has produced an updated version of space syntax which combines network measures common in the social sciences with visibility graphs to produce predictive models of movement within a set of three ancient Roman houses in Pompeii. These predictive models are tested through the navigation of virtual models of the houses by human subjects, demonstrating the significance of decoration in shaping movement, and, through quantitative and qualitative data, the value of digitally embodied phenomenology. This points ahead to the use of crowd-sourced, web-based global testing, diversifying the subject pool far beyond the narrow bounds of professional classicists or archaeologists.

show abstract

Section: Introductionmentioning

confidence: 99%