Ma, YajunKarim, AsifHaque, A. S. M. Farhan AlPonchio, FedericoPintus, Ruggero2022-09-262022-09-262022978-3-03868-178-62312-6124https://doi.org/10.2312/gch.20221217https://diglib.eg.org:443/handle/10.2312/gch20221217Acquiring knowledge about indigenous artefacts is difficult without prior experience. Due to the scarcity of digitally preserved artefacts, the new online generations are not exposed to even the most basic concepts. This paper demonstrates the feasibility of Image recognition technology for providing an app to describe aboriginal artefacts. By using Microsoft Custom Vision Service to classify and train the data, the predictive API of the model is called from a progressive web app developed to take photos of artefacts and retrieve their classification and description. We have divided aboriginal artefacts into the categories of tools or paintings for detection and description by machine learning systems. We test the trained model by taking new photos of artefacts already known to the system with the app camera. This performance was tested and compared with human classification to determine its usability. The results of the testing of our app show that the identification of aboriginal artefacts is applicable in certain cases.Attribution 4.0 International LicenseCCS Concepts: Applied Computing --> Computers in other domains; Digital libraries and archivesApplied ComputingComputers in other domainsDigital libraries and archives10.2312/gch.2022121715-184 pages