# Volumetric Attribute Compression for 3D Point Clouds using Feedforward Network with Geometric Attention

arXiv: https://arxiv.org/abs/2304.00335 · HTML: https://mlsanity.com/paper/2304.00335/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Builds on 23 of its 42 references that are in the map.
- First posted on arXiv 1 Apr 2023.

## Builds on (most cited first, 23 of 23)

- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [Variational image compression with a scale hyperprior](https://mlsanity.com/paper/1802.01436/) (2018)
- [End-to-end Optimized Image Compression](https://mlsanity.com/paper/1611.01704/) (2016)
- [Joint Autoregressive and Hierarchical Priors for Learned Image Compression](https://mlsanity.com/paper/1809.02736/) (2018)
- [High-Fidelity Generative Image Compression](https://mlsanity.com/paper/2006.09965/) (2020)
- [Full Resolution Image Compression with Recurrent Neural Networks](https://mlsanity.com/paper/1608.05148/) (2016)
- [Variable Rate Image Compression with Recurrent Neural Networks](https://mlsanity.com/paper/1511.06085/) (2015)
- [Nonlinear Transform Coding](https://mlsanity.com/paper/2007.03034/) (2020)
- [Variable Bitrate Neural Fields](https://mlsanity.com/paper/2206.07707/) (2022)
- [End-to-end optimization of nonlinear transform codes for perceptual quality](https://mlsanity.com/paper/1607.05006/) (2016)
- [Learning End-to-End Lossy Image Compression: A Benchmark](https://mlsanity.com/paper/2002.03711/) (2020)
- [Learning Convolutional Transforms for Lossy Point Cloud Geometry Compression](https://mlsanity.com/paper/1903.08548/) (2019)
- [Improved Deep Point Cloud Geometry Compression](https://mlsanity.com/paper/2006.09043/) (2020)
- [Graph Spectral Image Processing](https://mlsanity.com/paper/1801.04749/) (2018)
- [Deep AutoEncoder-based Lossy Geometry Compression for Point Clouds](https://mlsanity.com/paper/1905.03691/) (2019)
- [Deep Implicit Volume Compression](https://mlsanity.com/paper/2005.08877/) (2020)
- [Sparse Tensor-based Point Cloud Attribute Compression](https://mlsanity.com/paper/2204.01023/) (2022)
- [3D Scene Compression through Entropy Penalized Neural Representation Functions](https://mlsanity.com/paper/2104.12456/) (2021)
- [Efficient Nonlinear Transforms for Lossy Image Compression](https://mlsanity.com/paper/1802.00847/) (2018)
- [LVAC: Learned Volumetric Attribute Compression for Point Clouds using Coordinate Based Networks](https://mlsanity.com/paper/2111.08988/) (2021)
- [Graph-based compression of dynamic 3D point cloud sequences](https://mlsanity.com/paper/1506.06096/) (2015)
- [A Volumetric Approach to Point Cloud Compression](https://mlsanity.com/paper/1810.00484/) (2018)
- [Neural 3D Scene Compression via Model Compression](https://mlsanity.com/paper/2105.03120/) (2021)

## Built on by (most cited first, 4 of 4)

- [Volumetric 3D Point Cloud Attribute Compression: Learned polynomial bilateral filter for prediction](https://mlsanity.com/paper/2311.13533/) (2023)
- [Learned Nonlinear Predictor for Critically Sampled 3D Point Cloud Attribute Compression](https://mlsanity.com/paper/2311.13539/) (2023)
- [Deep Unrolling of Sparsity-Induced RDO for 3D Point Cloud Attribute Coding](https://mlsanity.com/paper/2509.08685/) (2025)
- [Constructing an Interpretable Deep Denoiser by Unrolling Graph Laplacian Regularizer](https://mlsanity.com/paper/2409.06676/) (2024)

## Abstract (arXiv)

We study 3D point cloud attribute compression using a volumetric approach: given a target volumetric attribute function f : ℝ3 rightarrow ℝ, we quantize and encode parameter vector θ that characterizes f at the encoder, for reconstruction f_hatθ(x) at known 3D points x's at the decoder. Extending a previous work Region Adaptive Hierarchical Transform (RAHT) that employs piecewise constant functions to span a nested sequence of function spaces, we propose a feedforward linear network that implements higher-order B-spline bases spanning function spaces without eigen-decomposition. Feedforward network architecture means that the system is amenable to end-to-end neural learning. The key to our network is space-varying convolution, similar to a graph operator, whose weights are computed from the known 3D geometry for normalization. We show that the number of layers in the normalization at the encoder is equivalent to the number of terms in a matrix inverse Taylor series. Experimental results on real-world 3D point clouds show up to 2-3 dB gain over RAHT in energy compaction and 20-30% bitrate reduction.