Self-Evolving Visual Concept Library using Vision-Language Critics

VLM
Re-assign community