[WACV 2025] LatteCLIP: Unsupervised CLIP Fine-Tuning via LMM-Synthetic Texts
Primary LanguageJupyter NotebookOtherNOASSERTION