Splannequin: Freezing Monocular Mannequin-Challenge Footage with Dual-Detection Splatting

TL;DR

Splannequin introduces a novel regularization technique for dynamic Gaussian splatting that improves the quality of frozen monocular scene reconstructions by effectively handling artifacts like ghosting and blur.

Contribution

It proposes an architecture-agnostic regularization method that detects hidden and defective Gaussian states, enhancing frozen scene synthesis from monocular videos without changing existing models.

Findings

Achieves 96% user preference for visual quality

Effectively reduces ghosting and blur artifacts

Requires no architectural changes or additional inference overhead

Abstract

Synthesizing high-fidelity frozen 3D scenes from monocular Mannequin-Challenge (MC) videos is a unique problem distinct from standard dynamic scene reconstruction. Instead of focusing on modeling motion, our goal is to create a frozen scene while strategically preserving subtle dynamics to enable user-controlled instant selection. To achieve this, we introduce a novel application of dynamic Gaussian splatting: the scene is modeled dynamically, which retains nearby temporal variation, and a static scene is rendered by fixing the model's time parameter. However, under this usage, monocular capture with sparse temporal supervision introduces artifacts like ghosting and blur for Gaussians that become unobserved or occluded at weakly supervised timestamps. We propose Splannequin, an architecture-agnostic regularization that detects two states of Gaussian primitives, hidden and defective, and applies temporal anchoring. Under predominantly forward camera motion, hidden states are anchored to their recent well-observed past states, while defective states are anchored to future states with stronger supervision. Our method integrates into existing dynamic Gaussian pipelines via simple loss terms, requires no architectural changes, and adds zero inference overhead. This results in markedly improved visual quality, enabling high-fidelity, user-selectable frozen-time renderings, validated by a 96% user preference. Project page: https://chien90190.github.io/splannequin/