OSA-LCM distills a portrait video diffusion model into a single-step generator that matches the quality of a 20-step teacher on FID/FVD, enabling near real-time talking-head generation.
Out of time: auto- mated lip sync in the wild
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
-
Real-time One-Step Diffusion-based Expressive Portrait Videos Generation
OSA-LCM distills a portrait video diffusion model into a single-step generator that matches the quality of a 20-step teacher on FID/FVD, enabling near real-time talking-head generation.