Increasing SAM Zero-Shot Performance on Multimodal Medical Images Using GPT-4 Generated Descriptive Prompts Without Human Annotation

Open in new window