Amortized Anchor Refinement for Deployable Continuous-Time 4D Gaussian Reconstruction
2026-08-31 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors address the challenge of creating detailed 4D reconstructions (3D shapes changing over time) on standalone XR headsets, which usually need too much computation. They introduce Amortized Anchor Refinement, a method that quickly predicts a rough model and then fine-tunes it a little within a fixed compute budget to keep details. They also apply a technique to clean up unstable parts while keeping important shapes intact, allowing smooth scene flow streaming. Their method works well on a benchmark and runs efficiently on typical consumer hardware and XR devices.
4D reconstructionXR headsetsGaussian representationAmortized Anchor Refinementscene flowpersistent homologyfeed-forward predictionoptimizationconsumer GPUStage-Capture benchmark
Authors
Jingong Chen, Qingwen Zhang, Sanghyeon Jun, Chulwoo Pack, Kyle Gao, Kwanghee Won
Abstract
Continuous-time 4D reconstruction remains impractical on standalone XR headsets. Per-scene optimization demands deployment-infeasible compute, and lower budgets cause collapse rather than degrade gradually. Feed-forward prediction is fast, but struggle to recover scene-specific detail. We present Amortized Anchor Refinement, which uses a frozen backbone to predict an initial Gaussian representation and a short optimization to specialize it under a fixed compute budget, with a capacity floor preserving representational density. A training-free stage then applies a persistent-homology constraint to prune unstable Gaussians while preserving topologically persistent structures, and streams the resulting trajectories directly as scene flow. On the Stage-Capture benchmark, Amortized Anchor Refinement achieves 24.31$\pm$2.22dB, while our deployment experiments demonstrate reconstruction within the target budget on a single consumer GPU and playback on a standalone XR headset.