This study introduces VADB, the largest video aesthetic database with 10,490 videos, and VADB-Net, a novel framework that outperforms existing models in video aesthetic assessment.