_en_search_text
_en_search_text
The HikarlA project
France 2030 award winner
Artificial intelligence algorithms will be trained to identify visual elements in a vast corpus produced between the mid-19th and early 20th centuries, from the Dubois collection (over 17 000 photographs), other Musée Guimet Japanese collections and similar collections accessible online (IIIF). These algorithms will be implemented on an analysis hub that will draw from the corpus new statistics on colour ranges, sequences of images, iconographic constants and novelties (duplicates, places, sets, accessories, models, typologies).
Photo (C) MNAAG, Paris, Dist. RMN-Grand Palais / image musée Guimet
Portrait d'une jeune fille japonaise 1877 – 1880, prise de vue, studio Stillfried & Andersen
The data gathered will allow to attribute ensembles by association and achieve a better understanding of the economy of studios, the aesthetic evolution of these productions and their intrinsic circulations (painters, models, photographs). These lessons will allow to question more broadly the spreading of images, exchanges between Europe and Eastern Asia, stereotypes (cultural, class, genre) and touristic practices at the dawn of globalisation. So the project will open up new horizons for research at the convergence of historical, aesthetic, economic and social challenges.
Extended over three years as of October 2023, the aim of the project is to spread as broadly as possible the results of the study through the platform (freely accessible), conferences, publications, and an exhibition.
An operation backed by the State in the context of the France 2030 device “Digitisation of the patrimony and architecture” run by the Caisse des Dépôts.
