A novel network for classification of cuneiform tablet metadata

Researchers developed a novel neural network architecture for classifying cuneiform tablet metadata using point-cloud representations. The convolution-inspired approach outperformed existing transformer-based methods like Point-BERT by gradually down-scaling point clouds while integrating local and global information.

• →New AI network architecture successfully classifies ancient cuneiform tablet metadata from high-resolution point-cloud data.
• →The method addresses challenges of limited annotated datasets in archaeological research through innovative down-scaling techniques.
• →Performance exceeds state-of-the-art transformer-based Point-BERT network in classification tasks.
• →Research addresses practical need as cuneiform tablet corpus size far exceeds available expert analysis capacity.
• →Source code and datasets will be made publicly available upon publication.