Skip to content
Book Open access

Filling the Unseen: Holistic Scene Extrapolation via 3D Gaussian Splatting

Sep 2026 · Proceedings of the 34th ACM International Conference on Multimedia · 0 citations · 38 references
Computer Science

TL;DR

This work designs an independent camera view detection mechanism to enable parallel conflict-free extrapolation, circumventing the reliance on error-prone cycle, and proposes a plug-and-play Quality-Aware Mask (QA-Mask) module, enabling selective utilization on generated data.

Abstract

3D Gaussian Splatting achieves photorealistic reconstruction within training view distribution, yet it degrades on out-of-distribution novel views, exhibiting holes in unobserved regions and artifacts in observable areas. Recent works formulat this task as extrapolation and interpolation and tried to address it with generative models, but remain limited in extrapolation scale and quality. They repeat a generate–reconstruct–shift cycle to progressively build a scene, which introduces accumulated errors with every step conditioning on previous outcomes. In this work, we propose a holistic framework for extrapolation and interpolation. We devise an independent camera view detection mechanism to enable parallel conflict-free extrapolation, circumventing the reliance on aforementioned error-prone cycle. Building upon this, we design a hierarchical pipeline that extrapolates independent and dependent camera views separately. Additionally, previous methods overlook inconsistency between generated and original images, resulting in compromising well-reconstructed areas. We propose a plug-and-play Quality-Aware Mask (QA-Mask) module, enabling selective utilization on generated data. By calibrating learning weights with pixel-wise rendering quality, it prevents generation-induced degradations on well-constructed areas. Extensive experiments demonstrate the superior performance of our framework, with QA-Mask generalizing on multiple generative reconstruction models (project page).

Read PDF

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.