Rethinking Human Evaluation Protocol for Text-to-Video Models: Enhancing Reliability,Reproducibility, and Practicality
This addresses reproducibility and practicality issues in human evaluation for text-to-video generation, which is incremental as it standardizes existing manual methods.
The paper tackles the problem of unreliable and costly manual evaluation for text-to-video models by introducing the T2VHE protocol, which reduces evaluation costs by nearly 50% while ensuring high-quality annotations.
Recent text-to-video (T2V) technology advancements, as demonstrated by models such as Gen2, Pika, and Sora, have significantly broadened its applicability and popularity. Despite these strides, evaluating these models poses substantial challenges. Primarily, due to the limitations inherent in automatic metrics, manual evaluation is often considered a superior method for assessing T2V generation. However, existing manual evaluation protocols face reproducibility, reliability, and practicality issues. To address these challenges, this paper introduces the Text-to-Video Human Evaluation (T2VHE) protocol, a comprehensive and standardized protocol for T2V models. The T2VHE protocol includes well-defined metrics, thorough annotator training, and an effective dynamic evaluation module. Experimental results demonstrate that this protocol not only ensures high-quality annotations but can also reduce evaluation costs by nearly 50\%. We will open-source the entire setup of the T2VHE protocol, including the complete protocol workflow, the dynamic evaluation component details, and the annotation interface code. This will help communities establish more sophisticated human assessment protocols.