InfLVG uses a GRPO-optimized context selection policy to choose top-K relevant video tokens for consistent, prompt-aligned long video generation.
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
-
InfLVG: Reinforce Inference-Time Consistent Long Video Generation with GRPO
InfLVG uses a GRPO-optimized context selection policy to choose top-K relevant video tokens for consistent, prompt-aligned long video generation.