By making temporal attention causal and sharing cached key-value features across all denoising steps, Ca2-VDM cuts autoregressive video generation cost from quadratic to roughly linear in the number of generated chunks.
(6) Since q and pθ are both Gaussian, DKL is determined by the mean µθ and covariance Σθ
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Ca2-VDM: Efficient Autoregressive Video Diffusion Model with Causal Generation and Cache Sharing
By making temporal attention causal and sharing cached key-value features across all denoising steps, Ca2-VDM cuts autoregressive video generation cost from quadratic to roughly linear in the number of generated chunks.