SIEVE: Multimodal Dataset Pruning using Image-Captioning Models (CVPR 2024)
Primary LanguageJupyter NotebookMIT LicenseMIT