Learnable Multi-level Discrete Wavelet Transforms for 3D Gaussian Splatting Frequency Modulation

2026-02-15

AI Summary

This work addresses the excessive growth of Gaussian primitives in 3D Gaussian Splatting, which arises from overfitting to fine details during training and leads to substantial memory and storage costs. To mitigate this issue, the authors propose a frequency modulation framework based on a multi-level learnable discrete wavelet transform. The approach recursively decomposes low-frequency subbands to construct a coarse-to-fine progressive supervision curriculum, effectively curbing the proliferation of Gaussians. Furthermore, it replaces the full high-pass filter with a single learnable scaling parameter, thereby alleviating gradient competition. Evaluated on standard benchmarks, the method significantly reduces the number of Gaussians while maintaining competitive rendering quality.

Technology Category

• Computer Vision: 3D Computer Vision
• Machine Learning: Learning with Manifolds
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

3D Gaussian Splatting (3DGS) has emerged as a powerful approach for novel view synthesis. However, the number of Gaussian primitives often grows substantially during training as finer scene details are reconstructed, leading to increased memory and storage costs. Recent coarse-to-fine strategies regulate Gaussian growth by modulating the frequency content of the ground-truth images. In particular, AutoOpti3DGS employs the learnable Discrete Wavelet Transform (DWT) to enable data-adaptive frequency modulation. Nevertheless, its modulation depth is limited by the 1-level DWT, and jointly optimizing wavelet regularization with 3D reconstruction introduces gradient competition that promotes excessive Gaussian densification. In this paper, we propose a multi-level DWT-based frequency modulation framework for 3DGS. By recursively decomposing the low-frequency subband, we construct a deeper curriculum that provides progressively coarser supervision during early training, consistently reducing Gaussian counts. Furthermore, we show that the modulation can be performed using only a single scaling parameter, rather than learning the full 2-tap high-pass filter. Experimental results on standard benchmarks demonstrate that our method further reduces Gaussian counts while maintaining competitive rendering quality.

Problem

• 3D Gaussian Splatting
• frequency modulation
• memory cost
• Gaussian densification
• novel view synthesis

Innovation

• Multi-level DWT
• 3D Gaussian Splatting
• Frequency Modulation
• Learnable Wavelet
• Curriculum Learning