Current text-to-video models struggle to answer real user queries that require video responses, scoring under 0.3 out of 1 on completeness in the new RealVideoQuest benchmark.
Coherence This metric measures whether the development process or steps of the response video content are logical and consistent and whether the consistency is reasonable
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Respond Beyond Language: A Benchmark for Video Generation in Response to Realistic User Intents
Current text-to-video models struggle to answer real user queries that require video responses, scoring under 0.3 out of 1 on completeness in the new RealVideoQuest benchmark.