PCA prunes video frames before encoding using similarity-based keyframe selection and temporally weighted aggregation, achieving about 2× speedup of VLLMs with minimal accuracy loss.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
PCA: Persistence-Aware Compression and Aggregation for Fast Video Large Language Models
PCA prunes video frames before encoding using similarity-based keyframe selection and temporally weighted aggregation, achieving about 2× speedup of VLLMs with minimal accuracy loss.