Improving Continual Learning for Gaussian Splatting based Environments Reconstruction on Commercial Off-the-Shelf Edge Devices

TL;DR

This paper introduces a precision-adaptive optimization framework that enables variational Bayesian Gaussian Splatting training on resource-constrained edge devices, significantly reducing memory and training time while maintaining high-quality 3D scene reconstruction.

Contribution

It presents a novel optimization approach that adapts precision levels and fuses kernels, allowing VBGS to run efficiently on embedded hardware without changing its core formulation.

Findings

Memory usage reduced from 9.44 GB to 1.11 GB.

Training time decreased from 234 min to 61 min.

Achieved 19x latency reduction on Jetson Orin Nano.

Abstract

Novel view synthesis (NVS) is increasingly relevant for edge robotics, where compact and incrementally updatable 3D scene models are needed for SLAM, navigation, and inspection under tight memory and latency budgets. Variational Bayesian Gaussian Splatting (VBGS) enables replay-free continual updates for the 3DGS algorithm by maintaining a probabilistic scene model, but its high-precision computations and large intermediate tensors make on-device training impractical. We present a precision-adaptive optimization framework that enables VBGS training on resource-constrained hardware without altering its variational formulation. We (i) profile VBGS to identify memory/latency hotspots, (ii) fuse memory-dominant kernels to reduce materialized intermediate tensors, and (iii) automatically assign operation-level precisions via a mixed-precision search with bounded relative error. Across the Blender, Habitat, and Replica datasets, our optimised pipeline reduces peak memory from 9.44 GB to 1.11 GB and training time from ~234 min to ~61 min on an A5000 GPU, while preserving (and in some cases improving) reconstruction quality of the state-of-the-art VBGS baseline. We also enable for the first time NVS training on a commercial embedded platform, the Jetson Orin Nano, reducing per-frame latency by 19x compared to 3DGS.