Year: 2025
Type: preprint
Abstract: Multi-view diffusion models have shown promise in 3D novel view synthesis, but most existing methods adopt a non-autoregressive formulation. This limits their applicability in world modeling, as they only support a fixed number of views and suffer from slow inference due to denoising all frames simultaneously. To address these limitations, we propose CausNVS, a multi-view diffusion model in an autoregressive setting, which supports arbitrary input-output view configurations and generates views sequentially. We train CausNVS with causal masking and per-frame noise, using pairwise-relative camera pose encodings (CaPE) for precise camera control. At inference time, we combine a spatially-aware sliding-window with key-value caching and noise conditioning augmentation to mitigate drift. Our experiments demonstrate that CausNVS supports a broad range of camera trajectories, enables flexible autoregressive novel view synthesis, and achieves consistently strong visual quality across diverse settings. Project page: https://kxhit.github.io/CausNVS.html.
Source: arXiv (Cornell University)
Authors: Kong, Xin, Watson, Daniel, Strümpler, Yannick, Niemeyer, Michael, Tombari, Federico
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W4415056089
Cites:
Cited by:
Related to: