Towards Zero-Shot Annotation of the Built Environment with Vision-Language Models (Vision Paper)

Open in new window