Reconstruct, Inpaint, Test-Time Finetune: Dynamic Novel-view Synthesis from Monocular Videos

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Chen, Kaihua, Khurana, Tarasha, Ramanan, Deva
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866911370752032768
author Chen, Kaihua
Khurana, Tarasha
Ramanan, Deva
author_facet Chen, Kaihua
Khurana, Tarasha
Ramanan, Deva
contents We explore novel-view synthesis for dynamic scenes from monocular videos. Prior approaches rely on costly test-time optimization of 4D representations or do not preserve scene geometry when trained in a feed-forward manner. Our approach is based on three key insights: (1) covisible pixels (that are visible in both the input and target views) can be rendered by first reconstructing the dynamic 3D scene and rendering the reconstruction from the novel-views and (2) hidden pixels in novel views can be "inpainted" with feed-forward 2D video diffusion models. Notably, our video inpainting diffusion model (CogNVS) can be self-supervised from 2D videos, allowing us to train it on a large corpus of in-the-wild videos. This in turn allows for (3) CogNVS to be applied zero-shot to novel test videos via test-time finetuning. We empirically verify that CogNVS outperforms almost all prior art for novel-view synthesis of dynamic scenes from monocular videos.
format Preprint
id arxiv_https___arxiv_org_abs_2507_12646
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Reconstruct, Inpaint, Test-Time Finetune: Dynamic Novel-view Synthesis from Monocular Videos
Chen, Kaihua
Khurana, Tarasha
Ramanan, Deva
Computer Vision and Pattern Recognition
We explore novel-view synthesis for dynamic scenes from monocular videos. Prior approaches rely on costly test-time optimization of 4D representations or do not preserve scene geometry when trained in a feed-forward manner. Our approach is based on three key insights: (1) covisible pixels (that are visible in both the input and target views) can be rendered by first reconstructing the dynamic 3D scene and rendering the reconstruction from the novel-views and (2) hidden pixels in novel views can be "inpainted" with feed-forward 2D video diffusion models. Notably, our video inpainting diffusion model (CogNVS) can be self-supervised from 2D videos, allowing us to train it on a large corpus of in-the-wild videos. This in turn allows for (3) CogNVS to be applied zero-shot to novel test videos via test-time finetuning. We empirically verify that CogNVS outperforms almost all prior art for novel-view synthesis of dynamic scenes from monocular videos.
title Reconstruct, Inpaint, Test-Time Finetune: Dynamic Novel-view Synthesis from Monocular Videos
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2507.12646