Gla-AI4BioMed at RRG24: Visual Instruction-tuned Adaptation for Radiology Report Generation
A visual language model leveraging CLIP and Vicuna-7B is fine-tuned to generate accurate radiology reports from chest X-rays, emphasizing the FINDINGS and IMPRESSIONS sections for improved medical interpretations.
Aug 1, 2024