Towards Label-free Scene Understanding by Vision Foundation Models

Open in new window