Answers are grounded in this paper and cite their sources.
The increasing availability of digital archaeological image collections has requested the need for advanced techniques in efficiently retrieving and analyzing artefacts. This research introduces a novel retrieval model designed to enhance the retrieval of archeological artefacts from large image databases. The proposed model leverages the hierarchical representations of pyramid features, capturing local and global visual information and then employs to selectively aggregate informative features for generating image embeddings. This not only enhances the discriminative power of the image descriptor but also facilitates the retrieval of artefacts across diverse scales and viewpoints. To capture both visual and semantic aspects, the network is trained by incorporating information related to cultures and time periods using a triplet-based algorithm. Experimental validation conducted on a dataset of historical artefacts from Thailand demonstrates the feasibility of the feature aggregation method, achieving a mean average precision of 80%, which is a significant improvement over existing deep learning-based image retrieval models.
No takes yet. Share an insight, caveat, or question.
Pipanmaekaporn et al. (2024) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: