Abstract: Multi-label image classification, which involves recognizing multiple objects within a single image, is a fundamental task in computer vision. Recently, Visual-Language Models (VLMs) have ...