Saved in:
Bibliographic Details
Main Authors: Lee, Jee Won, Choi, Jongseong Brad
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2511.02510
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908628069384192
author Lee, Jee Won
Choi, Jongseong Brad
author_facet Lee, Jee Won
Choi, Jongseong Brad
contents Sparse-voxel rasterization is a fast, differentiable alternative for optimization-based scene reconstruction, but it tends to underfit low-frequency content, depends on brittle pruning heuristics, and can overgrow in ways that inflate VRAM. We introduce LiteVoxel, a self-tuning training pipeline that makes SV rasterization both steadier and lighter. Our loss is made low-frequency aware via an inverse-Sobel reweighting with a mid-training gamma-ramp, shifting gradient budget to flat regions only after geometry stabilize. Adaptation replaces fixed thresholds with a depth-quantile pruning logic on maximum blending weight, stabilized by EMA-hysteresis guards and refines structure through ray-footprint-based, priority-driven subdivision under an explicit growth budget. Ablations and full-system results across Mip-NeRF 360 (6scenes) and Tanks & Temples (3scenes) datasets show mitigation of errors in low-frequency regions and boundary instability while keeping PSNR/SSIM, training time, and FPS comparable to a strong SVRaster pipeline. Crucially, LiteVoxel reduces peak VRAM by ~40%-60% and preserves low-frequency detail that prior setups miss, enabling more predictable, memory-efficient training without sacrificing perceptual quality.
format Preprint
id arxiv_https___arxiv_org_abs_2511_02510
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle LiteVoxel: Low-memory Intelligent Thresholding for Efficient Voxel Rasterization
Lee, Jee Won
Choi, Jongseong Brad
Computer Vision and Pattern Recognition
Sparse-voxel rasterization is a fast, differentiable alternative for optimization-based scene reconstruction, but it tends to underfit low-frequency content, depends on brittle pruning heuristics, and can overgrow in ways that inflate VRAM. We introduce LiteVoxel, a self-tuning training pipeline that makes SV rasterization both steadier and lighter. Our loss is made low-frequency aware via an inverse-Sobel reweighting with a mid-training gamma-ramp, shifting gradient budget to flat regions only after geometry stabilize. Adaptation replaces fixed thresholds with a depth-quantile pruning logic on maximum blending weight, stabilized by EMA-hysteresis guards and refines structure through ray-footprint-based, priority-driven subdivision under an explicit growth budget. Ablations and full-system results across Mip-NeRF 360 (6scenes) and Tanks & Temples (3scenes) datasets show mitigation of errors in low-frequency regions and boundary instability while keeping PSNR/SSIM, training time, and FPS comparable to a strong SVRaster pipeline. Crucially, LiteVoxel reduces peak VRAM by ~40%-60% and preserves low-frequency detail that prior setups miss, enabling more predictable, memory-efficient training without sacrificing perceptual quality.
title LiteVoxel: Low-memory Intelligent Thresholding for Efficient Voxel Rasterization
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2511.02510