AI-Enhanced Colour Reconstruction of Native Nanai Fish Skin Garments

A Community-Led Approach to Digital Heritage

Bibliographic Data

ID17823773
AuthorsElisa Palomino (0000-0002-4496-3701, Ca' Foscari University of Venice, corresponding author), Ana Cordoba Crespo (University of the Arts London), Anatol Donkan
Year2026
Volume41
Pagese00535-e00535
Publication date2026-04-01
Open AccessYes
TypeARTICLE
VenueDigital Applications in Archaeology and Cultural Heritage (JOURNAL)
Journal identifiersISSN: 2212-0548
PublisherElsevier BV (PUBLISHER)
DOI10.1016/j.daach.2026.e00535
OpenAlexW7160073095
LanguageEN
References cited19

Fish skin garments made by Native communities of the Lower Amur River in eastern Siberia were created by processing salmonid and carp species into wearable materials. These artefacts were dyed with pigments derived from local plants, minerals, and organic sources. Preserved in museum collections, their current condition reflects sometimes material degradation, discolouration, and past conservation treatments, limiting access to their original chromatic and surface properties. Digital reconstruction provides a method to examine these garments beyond their present state. While three-dimensional modelling and image processing have been applied in digital heritage, colour reconstruction remains limited for organic materials such as fish skin, particularly where documentation about provenance and processing methods is incomplete. This study reconstructs the colour and surface properties of a nineteenth century Nanai fish skin robe held at the Penn Museum. The methodology combines object-based research, historical dyeing reconstruction, high-resolution image sampling, AI-assisted 3D modelling, and CLO pattern-based garment simulation. Colourimetry supports the extraction of chromatic value of pigments, enabling accurate simulation of textures and tones. Ethnographic research informed the reconstruction of indigo, and alder bark pigments. The process was conducted in collaboration with a Nanai knowledge holder, whose input guided material interpretation. The study aims to reconstruct colour values from fish skin surfaces, simulate material behaviour through three-dimensional modelling, and integrate computational reconstruction with community knowledge.

Citizen science · Cultural heritage · Digital Imaging · Fish · Fish <Actinopterygii> · World heritage · Aesthetic Perception and Analysis · Generative Adversarial Networks and Image Synthesis · Music Technology and Sound Studies

Tools

Open DOIOpen Access