New ORGANIC3D model grounds 3D CT vision language models in organ-level image-text coupling
A paper introduces ORGANIC3D, a vision language model for 3D CT that grounds learning in organ-level image-text coupling, noting that high classification accuracy alone does not reflect model understanding.
Read the full story at arXiv cs.CV ↗
Timeline · 1 report
- 2026-11-19 05:00 · arXiv cs.CV
ORGANIC3D: Grounding 3D CT Vision Language Models in Organ Level Image Text Coupling