A cascaded training-free Video RAG pipeline decouples high-recall semantic prefetching from LLM-driven logical reranking to improve precision on cross-lingual long-video tasks with persona constraints.
arXiv preprint arXiv:2510.04428 , year=
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.CV 2years
2026 2representative citing papers
citing papers explorer
-
Decoupling Semantics and Logic: A Training-Free Coarse-to-Fine Pipeline for Video Retrieval-Augmented Generation
A cascaded training-free Video RAG pipeline decouples high-recall semantic prefetching from LLM-driven logical reranking to improve precision on cross-lingual long-video tasks with persona constraints.
- EgoMemReason: A Memory-Driven Reasoning Benchmark for Long-Horizon Egocentric Video Understanding