EvalCrafter: Benchmarking and Evaluating Large Video Generation Models

arXiv:2310.11440v3 Announce Type: replace
Abstract: The vision and language generative models have been overgrown in recent years. For video generation, various open-sourced models and public-available services have been developed to generate high-quality videos. However, these methods often use a few metrics, e.g., FVD or IS, to evaluate the performance. We argue that it is hard to judge the large conditional generative models from the simple metrics since these models are often trained on very large datasets with multi-aspect abilities. Thus, …

abstract arxiv benchmarking cs.cv generate generative generative models however language metrics performance public quality services type video video generation videos vision