FETV: A Benchmark for Fine-Grained Evaluation of Open-Domain Text-to-Video Generation

arXiv 2023 32.4 benchmark

TLDR

Introduces FETV, a multi-aspect, temporal-aware benchmark for fine-grained evaluation of open-domain text-to-video generation models and automatic metrics.

Reasoning

The paper addresses a clear gap in quantitative evaluation of text-to-video models with a well-structured benchmark and manual evaluations. Its main weakness is the lack of direct connection to world models or simulation, as the visible abstract focuses solely on video generation evaluation.

Read-first score

Read-first score 32.4, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 0.

Methodology quality 25%

Screens visible abstract and analysis fields for experiment, dataset, baseline, metric, and limitation evidence. markers=analysis,benchmark,evaluation,metric,result

Recency 8%

65.1

Uses a gentle age decay so recent papers surface without erasing older foundations. 2023

Reproducibility 25%

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=True; code=False; dataset=False; markers=github

Topical relevance 42%

Uses existing LLM keyword relevance scores normalized to 0-100. world model,world simulator,generative world model,interactive world model,video world model,world dynamics prediction,model-based reinforcement learning world model

Field roles

Methodology anchor

Rank sensitivity

Stability: volatile; rank range: 385.

Keyword Scores

world model

world simulator

generative world model

interactive world model

video world model

world dynamics prediction

model-based reinforcement learning world model

Resources

Paper