Training-Free Test-Time Adaptation via Shape and Style Guidance for Vision-Language Models

Open in new window