Ruixiang Xue

PhD Student

Vision Lab
Nanjing University (NJU)

Jiangsu Province, China

Profile picture

Introduction


I am a PhD student in the VISION lab of Nanjing University (NJU) and I am advised by Zhan Ma (NJU). I am mostly interested in Point Cloud Compression using deep learning approaches.

News


09-2024

Unicorn Part I:Geometry paper has been accepted to TPAMI!

09-2024

Unicorn Part II:Attribute paper has been accepted to TPAMI!

12-2023

NeRI paper has been accepted to ICASSP 2024!

11-2023

GR-Net paper has been accepted to TVCG 2023!

09-2023

I'm starting my PhD in NJU!

06-2023

I finish my B.S. in HDU!

Experience


02-2024 / 11-2024

Guangdong Oppo Mobile Telecommunications Corporation Ltd (OPPO)

Worked as a [Your Position] in the [Department/Team Name], focusing on [your key responsibilities]. Led and contributed to multiple projects related to [relevant technology/domain], optimizing [specific aspect] and improving [key metric or feature].

Key Contributions:

Publications


A Versatile Point Cloud Compressor Using Universal Multiscale Conditional Coding – Part I: Geometry

IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 2024

Jianqiang Wang#, Ruxiang Xue#, Jiaxin Li Dandan Ding Zhan Ma (#-Equally Contribution)
A universal multiscale conditional coding framework, Unicorn, is proposed to compress the geometry and attribute of any given point cloud. Geometry compression is addressed in Part I of this paper, while attribute compression is discussed in Part II. We construct the multiscale sparse tensors of each voxelized point cloud frame and properly leverage lower-scale priors in the current and (previously processed) temporal reference frames to improve the conditional probability approximation or content-aware predictive reconstruction of geometry occupancy in compression. Unicorn is a versatile, learning-based solution capable of compressing static and dynamic point clouds with diverse source characteristics in both lossy and lossless modes. Following the same evaluation criteria, Unicorn significantly outperforms standard-compliant approaches like MPEG G-PCC, V-PCC, and other learning-based solutions, yielding state-of-the-art compression efficiency while presenting affordable complexity for practical implementations.
  Paper   Code
                                @ARTICLE{10682566,
				  author={Wang, Jianqiang and Xue, Ruixiang and Li, Jiaxin and Ding, Dandan and Lin, Yi and Ma, Zhan},
				  journal={IEEE Transactions on Pattern Analysis and Machine Intelligence}, 
				  title={A Versatile Point Cloud Compressor Using Universal Multiscale Conditional Coding – Part II: Attribute}, 
				  year={2025},
				  volume={47},
				  number={1},
				  pages={252-268},
				  keywords={Geometry;Point cloud compression;Encoding;Tensors;Image color analysis;Image coding;Transform coding;Attribute;conditional coding;geometry;multiscale sparse representation;point cloud compression},
				  doi={10.1109/TPAMI.2024.3462945}}
                        

A Versatile Point Cloud Compressor Using Universal Multiscale Conditional Coding – Part II: Attribute

IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 2024

Jianqiang Wang, Ruxiang Xue, Jiaxin Li Dandan Ding Zhan Ma
A universal multiscale conditional coding framework, Unicorn, is proposed to code the geometry and attribute of any given point cloud. Attribute compression is discussed in Part II of this paper, while geometry compression is given in Part I of this paper. We first construct the multiscale sparse tensors of each voxelized point cloud attribute frame. Since attribute components exhibit very different intrinsic characteristics from the geometry element, e.g., 8-bit RGB color versus 1-bit occupancy, we process the attribute residual between lower-scale reconstruction and current-scale data. Similarly, we leverage spatially lower-scale priors in the current frame and (previously processed) temporal reference frame to improve the probability estimation of attribute intensity through conditional residual prediction in lossless mode or enhance the attribute reconstruction through progressive residual refinement in lossy mode for better performance. The proposed Unicorn is a versatile, learning-based solution capable of compressing a great variety of static and dynamic point clouds in both lossy and lossless modes. Following the same evaluation criteria, Unicorn significantly outperforms standard-compliant approaches like MPEG G-PCC, V-PCC, and other learning-based solutions, yielding state-of-the-art compression efficiency with affordable encoding/decoding runtime.
  Paper   Code
                                @ARTICLE{10682566,
				  author={Wang, Jianqiang and Xue, Ruixiang and Li, Jiaxin and Ding, Dandan and Lin, Yi and Ma, Zhan},
				  journal={IEEE Transactions on Pattern Analysis and Machine Intelligence}, 
				  title={A Versatile Point Cloud Compressor Using Universal Multiscale Conditional Coding – Part II: Attribute}, 
				  year={2025},
				  volume={47},
				  number={1},
				  pages={252-268},
				  keywords={Geometry;Point cloud compression;Encoding;Tensors;Image color analysis;Image coding;Transform coding;Attribute;conditional coding;geometry;multiscale sparse representation;point cloud compression},
				  doi={10.1109/TPAMI.2024.3462945}}
                      

neri

NeRI: Implicit Neural Representation Of LiDAR Point Cloud Using Range Image Sequence

ICASSP 2024

Ruxiang Xue, Jiaxin Li Tong Chen Dandan Ding Xun Cao Zhan Ma
This paper proposes the NeRI, an implicit neural representa- tion (INR) based LiDAR point cloud compressor. In NeRI, we first transform a sequence of 3D LiDAR frames into a 2D range image sequence through range image projection over time. Then, we employ a neural network conditioned on the temporal frame index and associated LiDAR sensor pose to fit input range images as closely as possible. The optimized net- work parameters, which implicitly represent the input LiDAR data, are later lossily compressed. NeRI decoder is then initialized using decoded parameters to generate range images for reconstructing the 3D LiDAR sequence accordingly. Extensive experimental results demonstrate the significant su- periority of NeRI regarding the compression efficiency and decoding speed compared to state-of-the-art 2D and 3D compressors for LiDAR point cloud.
  Paper   Code
                                @INPROCEEDINGS{10446596,
				  author={Xue, Ruixiang and Li, Jiaxin and Chen, Tong and Ding, Dandan and Cao, Xun and Ma, Zhan},
				  booktitle={ICASSP 2024 - 2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)}, 
				  title={NeRI: Implicit Neural Representation of LiDAR Point Cloud Using Range Image Sequence}, 
				  year={2024},
				  volume={},
				  number={},
				  pages={8020-8024},
				  keywords={Point cloud compression;Laser radar;Image coding;Three-dimensional displays;Quantization (signal);Neural networks;Transforms;Point cloud geometry compression;LiDAR data;implicit neural representation;range image},
				  doi={10.1109/ICASSP48485.2024.10446596}}
                        

neri

GRNet: Geometry Restoration for G-PCC Compressed Point Clouds Using Auxiliary Density Signaling

TVCG 2023

Gexin Liu Ruxiang Xue, Jiaxin Li Dandan Ding Zhan Ma
The lossy Geometry-based Point Cloud Compression (G-PCC) inevitably impairs the geometry information of point clouds, which deteriorates the quality of experience (QoE) in reconstruction and/or misleads decisions in tasks such as classification. To tackle it, this work proposes GRNet for the geometry restoration of G-PCC compressed large-scale point clouds. By analyzing the content characteristics of original and G-PCC compressed point clouds, we attribute the G-PCC distortion to two key factors: point vanishing and point displacement. Visible impairments on a point cloud are usually dominated by an individual factor or superimposed by both factors, which are determined by the density of the original point cloud. To this end, we employ two different models for coordinate reconstruction, termed Coordinate Expansion and Coordinate Refinement, to attack the point vanishing and displacement, respectively. In addition, 4-byte auxiliary density information is signaled in the bitstream to assist the selection of Coordinate Expansion, Coordinate Refinement, or their combination. Before being fed into the coordinate reconstruction module, the G-PCC compressed point cloud is first processed by a Feature Analysis Module for multiscale information fusion, in which kNN-based Transformer is leveraged at each scale to adaptively characterize neighborhood geometric dynamics for effective restoration. Following the common test conditions recommended in the MPEG standardization committee, GRNet significantly improves the G-PCC anchor and remarkably outperforms state-of-the-art methods on a great variety of point clouds (e.g., solid, dense, and sparse samples) both quantitatively and qualitatively. Meanwhile, GRNet runs fairly fast and uses a smaller-size model when compared with existing learning-based approaches, making it attractive to industry practitioners.
  Paper   Code
                                @misc{guedon2023sugar,
                                    title={SuGaR: Surface-Aligned Gaussian Splatting for Efficient 3D Mesh Reconstruction and High-Quality Mesh Rendering}, 
                                    author={Antoine Guédon and Vincent Lepetit},
                                    year={2023},
                                    eprint={2311.12775},
                                    archivePrefix={arXiv},
                                    primaryClass={cs.GR}
                              }
                        

Preprint


neri

Efficient LiDAR Point Cloud Geometry Compression Through Neighborhood Point Attention

Ruxiang Xue, Jianqiang Wang Zhan Ma
Although convolutional representation of multiscale sparse tensor demonstrated its superior efficiency to accurately model the occupancy probability for the compression of geometry component of dense object point clouds, its capacity for representing sparse LiDAR point cloud geometry (PCG) was largely limited. This is because 1) fixed receptive field of the convolution cannot characterize extremely and unevenly distributed sparse LiDAR points very well; and 2) pretrained convolutions with fixed weights are insufficient to dynamically capture information conditioned on the input. This work therefore suggests the neighborhood point attention (NPA) to tackle them, where we first use k nearest neighbors (kNN) to construct adaptive local neighborhood; and then leverage the self-attention mechanism to dynamically aggregate information within this neighborhood. Such NPA is devised as a NPAFormer to best exploit cross-scale and same-scale correlations for geometric occupancy probability estimation. Compared with the anchor using standardized G-PCC, our method provides >17% BD-rate gains for lossy compres- sion, and >14% bitrate reduction for lossless scenario using popular LiDAR point clouds in SemanticKITTI and Ford datasets. Compared with the state-of-the-art (SOTA) solution using attention optimized octree coding method, our approach requires much less decoding runtime with about 640× speedup on average, while still presenting better compression efficiency.
  Paper   Code
                                @article{xue2022efficient,
				  title={Efficient lidar point cloud geometry compression through neighborhood point attention},
				  author={Xue, Ruixiang and Wang, Jianqiang and Ma, Zhan},
				  journal={arXiv preprint arXiv:2208.12573},
				  year={2022}
				}
                        

© You are welcome to copy the code, please attribute the source with a link back to this page.
Template inspired from Tom Monnier and Mathis Petrovich.