AI RESEARCH

ReCoSplat: Autoregressive Feed-Forward Gaussian Splatting Using Render-and-Compare

arXiv CS.CV

ArXi:2603.09968v1 Announce Type: new Online novel view synthesis remains challenging, requiring robust scene reconstruction from sequential, often unposed, observations. We present ReCoSplat, an autoregressive feed-forward Gaussian Splatting model ing posed or unposed inputs, with or without camera intrinsics. While assembling local Gaussians using camera poses scales better than canonical-space prediction, it creates a dilemma during