A cross-view attention module and hybrid training recipe let a frozen text-to-video model generate synchronized multi-camera videos from arbitrary viewpoints given only a text prompt and camera poses.
Generative novel view synthesis with 3d-aware diffusion models
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
-
SynCamMaster: Synchronizing Multi-Camera Video Generation from Diverse Viewpoints
A cross-view attention module and hybrid training recipe let a frozen text-to-video model generate synchronized multi-camera videos from arbitrary viewpoints given only a text prompt and camera poses.