CoralLite uses deep learning to reconstruct coral skeletons

Researchers have developed CoralLite, a new deep learning model and dataset for reconstructing coral colonies from micro-CT scans. The model, a hybrid V-Trans-UNet architecture, is trained on detailed annotations to identify and segment individual corallites, which archive a coral's life history. CoralLite aims to help scientists understand skeletal growth rates and polyp division by tracking these structures, demonstrating the potential of machine learning in marine biology research. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT Enables detailed 3D modeling of coral skeletons, advancing marine biology research and conservation efforts.

RANK_REASON The cluster contains an academic paper detailing a new method and dataset for a specific scientific application. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• CoralLite
• V-Trans-UNet
• arXiv

• paper
• other