CP-LLM uses dual vision encoders in a multimodal LLM to separately handle video context and pixel distortions, then reasons about both to output quality scores and descriptions with claimed SOTA cross-dataset performance.
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
UNVERDICTED 1representative citing papers
citing papers explorer
-
Context and Pixel Aware Large Language Model for Video Quality Assessment
CP-LLM uses dual vision encoders in a multimodal LLM to separately handle video context and pixel distortions, then reasons about both to output quality scores and descriptions with claimed SOTA cross-dataset performance.