arXiv：cs.LG（机器学习，全量分类）· Haolong Wang, Yicheng Zhan, Kaan Ak\c{s}it, Simeng Qiu·· 14 小时前AI 评分33
CVQPG：二次相位高斯泼溅的全息图表示方法
Hologram Representation via Quadratic Phase Gaussian Splatting
AI 导读
研究者提出复数二次相位高斯（CVQPG），为每个 2D 高斯基元增加由可学习曲率参数控制的二次相位轮廓。在相同基元数量下，全息重建平均 PSNR 较平面高斯基线提升 0.19 dB（RGB）和 0.33 dB（灰度）；相同参数量下提升 0.05 dB（RGB）和 0.08 dB（灰度），并在所有视觉质量指标上保持领先。频域分析显示中高频段重建 MSE 最多下降 11%（RGB）和 22%（灰度）。
正文
Abstract:We introduce Complex-Valued Quadratic Phase Gaussian (CVQPG), a novel hologram representation method that augments each 2D Gaussian primitive with a quadratic phase profile controlled by a learnable curvature parameter. Against the planar Gaussian baseline, CVQPG improves the average PSNR of holographic reconstructions by 0.19 dB (RGB) and 0.33 dB (grayscale) at equal primitive counts, and by 0.05 dB (RGB) and 0.08 dB (grayscale) at equal parameter counts, where it still leads in all visual quality metrics. Our frequency-domain analysis shows that CVQPG better preserves the mid-to-high frequency band of natural images, where the reconstruction MSE drops by up to 11% (RGB) and 22% (grayscale), indicating that modulating primitive wavefronts is an effective and lightweight enhancement.
| Comments: | SIGGRAPH Asia 2026 Technical Communications |
| Subjects: | Graphics (cs.GR); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG) |
| Cite as: | arXiv:2609.11434 [cs.GR] |
| (or arXiv:2609.11434v2 [cs.GR] for this version) | |
| https://doi.org/10.48550/arXiv.2609.11434 arXiv-issued DOI via DataCite |
|
| Related DOI: | https://doi.org/10.1145/3829339.3847858
DOI(s) linking to related resources |
Submission history
From: Haolong Wang [view email]
[v1]
Thu, 10 Sep 2026 12:06:11 UTC (12,968 KB)
[v2]
Thu, 1 Oct 2026 17:08:38 UTC (9,686 KB)
来源：arXiv：cs.LG（机器学习，全量分类） · arxiv.org