Balinese Carving Ornaments Classification Using InceptionResnetV2 Architecture

Abstract

All types of Balinese carving ornaments have special categories and names, but not many people know and understand them. This research conducts a study to build a classification system and automatic identification of types of Balinese carving ornaments based on digital images using deep learning-based methods, namely InceptionResnetV2 architecture. This architecture is tested as a comparison with the previous reported research results using feature extraction-based methods and with classifiers based on neural networks and multilayer perceptrons. The experimental results show that the best accuracy values obtained using the InceptionResnetV2 architecture is 76.66%. This result will be very useful for the development of further methods and systems.