P-SWA enables parallel decoding via diagonal wavefronts, a hyperprior, and an accumulator, delivering 36% higher decoding speed than parallel VCT and BD-rate savings of up to 10.0% on I-frames and 7.1% on P-frames versus SWA.
Experimental setting For a fair comparison, we strictly adhere to the training and testing protocols established in [3]
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
eess.IV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Parallel Context Modeling for Sliding Window Attention in Neural Video Coding
P-SWA enables parallel decoding via diagonal wavefronts, a hyperprior, and an accumulator, delivering 36% higher decoding speed than parallel VCT and BD-rate savings of up to 10.0% on I-frames and 7.1% on P-frames versus SWA.