A Survey of Computational Techniques for Automated Video Creation and their Evaluation

This study evaluated the inter-rater agreement among five evaluators on automatic video generation applications using Text-2 Video Computation models. The evaluation involved 15 criteria and used statistical analysis techniques to determine agreement. The results showed moderate agreement, with the highest-scoring applications being App A3 (Pictory), App A5 (DeebrainAI), and App A10 (elai). The study contributes to understanding inter-rater agreement in evaluating automatic video generation applications and highlights the highest-rated ones.