SplineGS: Robust Motion-Adaptive Spline for Real-Time Dynamic 3D Gaussians from Monocular Video

Jongmin Park 1, Minh-Quan Viet Bui 1, Juan Luis Gonzalez Bello 1, Jaeho Moon 1, Munchurl Kim 1, 2 Chung-Ang University, [email protected] · 2024

Read SplineGS: Robust Motion-Adaptive Spline for Real-Time Dynamic 3D Gaussians from Monocular Video by Jongmin Park 1, Minh-Quan Viet Bui 1, Juan Luis Gonzalez Bello 1, Jaeho Moon 1, Munchurl Kim 1, 2 Chung-Ang University, [email protected] (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-09982.txt.

Description

Park and five co-authors rebuilt moving scenes from one camera in this paper from twenty twenty-four. A single phone video lacks the many views that 3D reconstruction usually needs. Their method, SplineGS, fills the scene with soft 3D blobs that drift along smooth spline paths. A few control points steer each path.

A pruning step drops unneeded control points while keeping the motion faithful. Camera poses and blob traits are tuned together, skipping the usual pre-processing pipeline. That design hardens the method against messy real-world footage. Tests beat leading rivals on view quality for dynamic scenes. Rendering ran thousands of times faster than the reconstruction methods it replaces.

Catalog record

CREATORJongmin Park 1, Minh-Quan Viet Bui 1, Juan Luis Gonzalez Bello 1, Jaeho Moon 1, Munchurl Kim 1, 2 Chung-Ang University, [email protected]
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE54 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-09982

Browse the full library · About the Tower of Babel