VidChain improves dense video captioning by decomposing the task into sub-tasks and using metric-based Direct Preference Optimization with a preference gap threshold.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
VidChain: Chain-of-Tasks with Metric-based Direct Preference Optimization for Dense Video Captioning
VidChain improves dense video captioning by decomposing the task into sub-tasks and using metric-based Direct Preference Optimization with a preference gap threshold.