pith. sign in

arxiv: 2605.18006 · v1 · pith:SCSG4DIDnew · submitted 2026-05-18 · 📡 eess.IV · cs.CV· cs.MM

Inter-LPCM: Learning-based Inter-Frame Predictive Coding for LiDAR Point Cloud Compression

Chang Sun , Hui Yuan , Shiqi Jiang , Chongzhen Tian , Guanghui Zhang , Raouf Hamzaoui This is my paper

Pith reviewed 2026-05-20 00:41 UTC · model grok-4.3

classification 📡 eess.IV cs.CVcs.MM
keywords LiDAR point cloud compressioninter-frame predictionspherical coordinateslearning-based codingrate-distortion optimizationgeometry compressionattention mechanism
0
0 comments X

The pith

A learning-based inter-frame predictor for radius and elevation in spherical coordinates improves rate-distortion performance for LiDAR point cloud compression over linear models.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper sets out to demonstrate that a learned inter-frame approach can reduce the bits required to represent LiDAR geometry data while maintaining reconstruction quality. It replaces the linear prediction used in the existing PredGeom standard with two new models: one that estimates each point's radius from neighbors in both the current frame and a registered reference frame, and another that uses lightweight attention to predict elevation by relating coordinates across the cloud. Azimuth is handled with simple delta coding based on the sensor's fixed angular resolution, quantization steps are chosen to optimize the rate-distortion trade-off, and separate entropy models are built for each spherical component to match their different statistics. If these changes work as intended, streaming or archival storage of vehicle and mapping sensor data could become more efficient without abandoning the convenient spherical parameterization that matches how LiDARs actually scan the world.

Core claim

The paper claims that an inter-frame radius predictive model, which draws on neighboring points from both the current and registered reference frames, together with a lightweight attention-based elevation predictor that captures long-range geometric correlations, allows more accurate estimation of point positions than the linear model in PredGeom. When combined with delta coding for azimuth, RD-optimized quantization in spherical coordinates, and coordinate-specific entropy models, the resulting Inter-LPCM method produces better rate-distortion curves on LiDAR sequences.

What carries the argument

The inter-frame radius predictive (Inter-RP) model that estimates the current point's radius from neighbors in both frames, plus the lightweight attention-based prediction (LAEP) model that relates elevation angles across coordinates.

If this is right

  • Lower bitrates are achieved for the same geometry distortion on standard LiDAR test sets.
  • Separate entropy models for azimuth, radius, and elevation yield tighter probability estimates than a single shared model.
  • RD-optimized quantization steps in spherical coordinates further reduce the number of bits needed per point.
  • The method preserves the fixed angular resolution structure of raw LiDAR scans while adding inter-frame redundancy removal.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same inter-frame neighborhood and attention idea could be adapted to compress other range-sensor data such as radar point clouds.
  • Hybrid systems that combine this predictive coding with octree or voxel-based methods might achieve still lower rates on static scenes.
  • Evaluating the attention mechanism on high-speed sequences with rapid object motion would test how well long-range correlations hold under strong temporal change.

Load-bearing premise

Accurate registration between the reference frame and the current frame is always available, and predictions based on neighboring points plus attention will not create new artifacts that raise the overall bitrate.

What would settle it

Compress a set of LiDAR sequences after introducing controlled registration errors of several centimeters and check whether the resulting rate-distortion performance drops below that of the baseline PredGeom coder.

Figures

Figures reproduced from arXiv: 2605.18006 by Chang Sun, Chongzhen Tian, Guanghui Zhang, Hui Yuan, Raouf Hamzaoui, Shiqi Jiang.

Figure 1
Figure 1. Figure 1: Acquisition of LPCs. First, LiDAR uses multiple laser emitters with [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Overall architecture of the proposed Inter-LPCM. First, the current point cloud [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 4
Figure 4. Figure 4: Architecture of the proposed entropy model. The encoded residuals [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Compression pipeline for radii. In the partitioning stage, both the current frame [PITH_FULL_IMAGE:figures/full_fig_p006_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Radius variance of predictive trees with different laser IDs [PITH_FULL_IMAGE:figures/full_fig_p006_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Architecture of the proposed LAEP model. The model consists of [PITH_FULL_IMAGE:figures/full_fig_p008_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: RD performance of the proposed method and the baselines on the [PITH_FULL_IMAGE:figures/full_fig_p010_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Visualization of point clouds reconstructed from OctAttention, SCP-EHEM, and the proposed Inter-LPCM on the [PITH_FULL_IMAGE:figures/full_fig_p011_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Rate–D1 PSNR curves of selecting Qs values using different numbers [PITH_FULL_IMAGE:figures/full_fig_p012_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Bitrates for compressing azimuth angles using the normal distribution [PITH_FULL_IMAGE:figures/full_fig_p012_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Rate–D1 PSNR curves obtained by forming P lower with thresholds τ = 0.2, 0.4, 0.8 and 1.6. models the probability distribution of azimuth angles, resulting in a lower bitrate. We also evaluated the effectiveness of each component in the encoder of Radius. As shown in Table VIII, removing the partitioning stage leads to degraded RD performance. This degradation arises from variations in ground points induc… view at source ↗
read the original abstract

Because LiDAR sensors acquire point clouds with a fixed angular resolution, the resulting data can be systematically parameterized and efficiently compressed in the spherical coordinate system. Traditional spherical coordinate-based point cloud compression methods have demonstrated strong rate-distortion (RD) performance, with the predictive geometry coding (PredGeom) method in the geometry-based point cloud compression (G-PCC) standard being a prominent example. Although PredGeom includes an inter-frame prediction mode, it relies on a simple linear model, which limits its ability to capture complex motion patterns and structural dependencies. Meanwhile, existing learning-based compression methods in the spherical domain do not exploit inter-frame correlations to reduce geometry redundancy. To address these limitations, we propose a learning-based inter-frame predictive coding method, termed Inter-LPCM. For azimuth prediction, we employ a delta coding strategy based on the predefined angular resolution. To improve radius compression, we introduce an inter-frame radius predictive (Inter-RP) model that estimates the current point's radius using neighboring points from both the current frame and the registered reference frame. In addition, we design a lightweight attention-based prediction (LAEP) model to predict elevation angles by capturing long-range geometric correlations across different coordinates. For quantization, we propose an RD-optimized method to select quantization steps in the spherical coordinate system. For entropy coding, we design distinct models for each spherical coordinate component. These models are adapted to the statistical priors of each coordinate, enabling more accurate probability estimation. Our source code is publicly available at https://github.com/SDUChangSun/Inter-LPCM

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript proposes Inter-LPCM, a learning-based inter-frame predictive coding method for LiDAR point cloud compression in spherical coordinates. It extends the PredGeom approach from G-PCC by replacing the linear inter-frame model with an inter-frame radius predictive (Inter-RP) model that estimates radius from neighboring points in both the current frame and a registered reference frame, a lightweight attention-based elevation predictor (LAEP) that captures long-range geometric correlations across coordinates, delta coding for azimuth, an RD-optimized quantization step selector, and coordinate-specific entropy coding models. The central claim is that these learned components better capture complex motion and structural dependencies, yielding improved rate-distortion performance over existing methods.

Significance. If the reported rate-distortion gains are robust, the work could advance efficient compression of dynamic LiDAR sequences for bandwidth-constrained applications such as autonomous driving. The public release of source code supports reproducibility and is a clear strength. However, the practical significance hinges on whether the inter-frame gains remain when registration is imperfect, as the method's advantage over PredGeom is predicated on the learned predictors receiving correctly aligned context.

major comments (2)
  1. [§3.2] §3.2 (Inter-RP model description): The radius prediction explicitly conditions on neighboring points from the registered reference frame, yet the manuscript provides no sensitivity analysis or ablation under controlled registration error (e.g., added ego-motion noise or scene dynamics). This is load-bearing for the central claim because misalignment would supply incorrect context to the learned predictor, potentially increasing quantized residuals and entropy beyond the linear PredGeom baseline.
  2. [§4] §4 (Experimental results): The claimed RD improvement is not accompanied by registration-error robustness tests or ablation isolating the contribution of accurate alignment versus the attention mechanism. Without these, it is unclear whether the reported gains are attributable to superior modeling or to the assumption of near-perfect registration that may not hold in real LiDAR streams.
minor comments (2)
  1. [Abstract] The abstract would be strengthened by including one or two key quantitative RD metrics (e.g., BD-rate savings versus PredGeom) rather than only qualitative statements.
  2. [§3.3] Notation for the attention weights and neighbor selection in LAEP could be made more explicit with a short equation or diagram to aid readability.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback on our manuscript. We have carefully reviewed the major comments and provide point-by-point responses below. We agree that additional robustness analysis would strengthen the claims and will incorporate the suggested experiments in the revised manuscript.

read point-by-point responses

  1. Referee: [§3.2] §3.2 (Inter-RP model description): The radius prediction explicitly conditions on neighboring points from the registered reference frame, yet the manuscript provides no sensitivity analysis or ablation under controlled registration error (e.g., added ego-motion noise or scene dynamics). This is load-bearing for the central claim because misalignment would supply incorrect context to the learned predictor, potentially increasing quantized residuals and entropy beyond the linear PredGeom baseline.

    Authors: We acknowledge that the performance of the Inter-RP model depends on the quality of the registration between frames. While the manuscript assumes accurate registration as is common in inter-frame coding literature, we agree that a sensitivity analysis is valuable. In the revised version, we will add experiments introducing controlled registration errors (e.g., noise in translation and rotation parameters) and evaluate the RD performance of Inter-LPCM compared to PredGeom under these conditions. This will provide insight into the robustness of the learned predictor. revision: yes

  2. Referee: [§4] §4 (Experimental results): The claimed RD improvement is not accompanied by registration-error robustness tests or ablation isolating the contribution of accurate alignment versus the attention mechanism. Without these, it is unclear whether the reported gains are attributable to superior modeling or to the assumption of near-perfect registration that may not hold in real LiDAR streams.

    Authors: We appreciate this observation. To isolate the contributions, we will perform additional ablations in Section 4: one varying the registration accuracy and another disabling the attention mechanism in LAEP while keeping other components fixed. These results will be included in the revised manuscript to clarify that the gains arise from the learned inter-frame models rather than solely from perfect alignment assumptions. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation introduces independent learned predictors

full rationale

The paper defines Inter-RP and LAEP as new neural models that take registered reference-frame neighbors as input to predict radius and elevation, then applies standard RD-optimized quantization and per-coordinate entropy models. None of these steps reduce by construction to quantities already fitted inside the paper's own equations, nor do they rely on self-citation chains or imported uniqueness theorems; the claimed RD gains are presented as arising from the added modeling capacity relative to the linear PredGeom baseline, which remains an external reference.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The approach rests on the domain assumption that LiDAR data admits a fixed-angular-resolution spherical parameterization and that inter-frame registration can be performed reliably; no new physical entities or ad-hoc constants are introduced beyond standard learned model weights.

axioms (1)
  • domain assumption LiDAR point clouds can be systematically parameterized in the spherical coordinate system due to fixed angular resolution.
    Stated in the opening paragraph of the abstract as the foundation for all subsequent compression steps.

pith-pipeline@v0.9.0 · 5831 in / 1234 out tokens · 41721 ms · 2026-05-20T00:41:40.624672+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

61 extracted references · 61 canonical work pages

  1. [1]

    Point Cloud Compression with Range Image- Based Entropy Model for Autonomous Driving,

    S. Wang and M. Liu, “Point Cloud Compression with Range Image- Based Entropy Model for Autonomous Driving,” inProc. of the European Conference on Computer Vision, Tel Aviv, Israel, 2022, pp. 323-340

    work page 2022

  2. [2]

    Efficient and Robust LiDAR-Based End-to-End Navigation,

    Z. Liu, A. Amini, S. Zhu, S. Karaman, S. Han and D. L. Rus, “Efficient and Robust LiDAR-Based End-to-End Navigation,” inProc. of the IEEE International Conference on Robotics and Automation, Xi’an, China, 2021, pp. 13247-13254

    work page 2021

  3. [3]

    Geographical Map Registration and Fusion of Lidar-Aerial Orthoimagery in GIS,

    S. Yi, S. Worrall and E. Nebot, “Geographical Map Registration and Fusion of Lidar-Aerial Orthoimagery in GIS,” inProc. of the IEEE Intelligent Transportation Systems Conference, Auckland, New Zealand, 2019, pp. 128-134

    work page 2019

  4. [4]

    SCP: Spherical-Coordinate-Based Learned Point Cloud Compression,

    A. Luo, L. Song, K. Nonaka, K. Unno, H. Sun, M. Goto, and J. Katto, “SCP: Spherical-Coordinate-Based Learned Point Cloud Compression,” inProc. of the AAAI Conference on Artificial Intelligence, Vancouver, Canada, 2024, pp. 3954-3962

    work page 2024

  5. [5]

    Efficient Hierarchical Entropy Model for Learned Point Cloud Compression,

    R. Song, C. Fu, S. Liu, and G. Li, “Efficient Hierarchical Entropy Model for Learned Point Cloud Compression,” inProc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Vancouver, Canada, 2023, pp. 14368-14377

    work page 2023

  6. [6]

    OctAttention: Octree-Based Large-Scale Contexts Model for Point Cloud Compression,

    C. Fu, G. Li, R. Song, W. Gao, and S. Liu, “OctAttention: Octree-Based Large-Scale Contexts Model for Point Cloud Compression,” inProc. of the AAAI Conference on Artificial Intelligence, online, 2022, pp. 625-633

    work page 2022

  7. [7]

    Predictive Geometry Coding,

    MPEG 3D Graphics Coding and Haptics Coding, “Predictive Geometry Coding,” inISO/IEC ITCI/SC29/WG1I MPEG output document m51012, Macau, Oct. 2019

    work page 2019

  8. [8]

    Predictive geom- etry angular mode using spherical LiDAR data input,

    MPEG 3D Graphics Coding and Haptics Coding, “Predictive geom- etry angular mode using spherical LiDAR data input,” inISO/IEC ITCI/SC29/WG1I MPEG output document m55361, Macau, Oct. 2020

    work page 2020

  9. [9]

    Emerging MPEG standards for point cloud compression,

    S. Schwarz, M. Preda, V . Baroncini, M. Budagavi, P. Cesar, P. A. Chou, R. A. Cohen, M. Krivoku´ca, S. Lasserre, Z. Li, J. Llach, K. Mammou, R. Mekuria, O. Nakagami, E. Siahaan, A. Tabatabai, A. M. Tourapis, and V . Zakharchenko, “Emerging MPEG standards for point cloud compression,” IEEE Journal on Emerging and Selected Topics in Circuits and Systems, vo...

    work page 2019

  10. [10]

    Octree-based Point-Cloud Compression,

    R. Schnabel and R. Klein, “Octree-based Point-Cloud Compression,” Symposium on Point Based Graphics, Boston, Massachusetts, USA, 2006, pp. 111-120

    work page 2006

  11. [11]

    C. Sun, H. Yuan, S. Jiang, D. Ai, W. Zhang, R. Hamzaoui, ”LPCM: Learning-based Predictive Coding for LiDAR Point Cloud Compression,” arXiv preprint arXiv:2505.20059

    work page arXiv

  12. [12]

    Sze and M

    V . Sze and M. Budagavi, ”High Throughput CABAC Entropy Coding in HEVC,”IEEE Transactions on Circuits and Systems for Video Tech- nology, vol. 22, no. 12, pp. 1778-1791, Dec. 2012

    work page 2012

  13. [13]

    Enhanced G-PCC Test Model v31,

    MPEG 3D Graphics Coding and Haptics Coding, “Enhanced G-PCC Test Model v31,”ISO/IEC JTC 1/SC 29/WG7 N01232 output document, Macau, Jul. 2025

    work page 2025

  14. [14]

    Enhanced G-PCC codec description,

    MPEG 3D Graphics Coding and Haptics Coding, “Enhanced G-PCC codec description,” inISO/IEC JTC1/SC29/WG07 MPEG output docu- ment N01014, Online, Nov. 2024

    work page 2024

  15. [15]

    Report on Triangle soup,

    MPEG 3D Graphics Coding and Haptics Coding, “Report on Triangle soup,”ISO/IEC JTC1/SC29/WG7 m58005, Macau, Oct. 2021

    work page 2021

  16. [16]

    Description of Explo- ration Experiment 13.67 on dynamic OBUF,

    MPEG 3D Graphics Coding and Haptics Coding, “Description of Explo- ration Experiment 13.67 on dynamic OBUF,”ISO/IECJTC1/SC29/WG7 MPEG input document N559, Macau, Jan. 2023

    work page 2023

  17. [17]

    Inter prediction with predictive geometry coding,

    MPEG 3D Graphics Coding and Haptics Coding, “Inter prediction with predictive geometry coding,”ISO/IEC/JTC1/SC29/WG7 m56117, Macau, jan. 2021

    work page 2021

  18. [18]

    G-PCC Performance Review,

    MPEG 3D Graphics Coding and Haptics Coding, “G-PCC Performance Review,”ISO/IEC JTC 1/SC29/WG7 MPEG input document M72237, Macau, Apr. 2025

    work page 2025

  19. [19]

    Krivoku ´ca, P

    M. Krivoku ´ca, P. A. Chou and M. Koroteev, ”A V olumetric Approach to Point Cloud Compression–Part II: Geometry Compression,”IEEE Transactions on Image Processing, vol. 29, pp. 2217-2229, 2020. JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 14

    work page 2020

  20. [20]

    D. E. O. Tzamarias, K. Chow, I. Blanes and J. Serra-Sagrist `a, ”Fast Run-Length Compression of Point Cloud Geometry,”IEEE Transactions on Image Processing, vol. 31, pp. 4490-4501, 2022

    work page 2022

  21. [21]

    D. Li, K. Ma, J. Wang and G. Li, ”Hierarchical Prior-Based Super Resolution for Point Cloud Geometry Compression,”IEEE Transactions on Image Processing, vol. 33, pp. 1965-1976, 2024

    work page 1965

  22. [22]

    A Regularized Projection- Based Geometry Compression Scheme for LiDAR Point Cloud,

    Y . Yu, W. Zhang, G. Li and F. Yang, “A Regularized Projection- Based Geometry Compression Scheme for LiDAR Point Cloud,”IEEE Transactions on Circuits and Systems for Video Technology, vol. 33, no. 3, pp. 1427-1437, Mar. 2023

    work page 2023

  23. [23]

    A Comprehensive Study and Comparison of Core Technologies for MPEG 3-D Point Cloud Compression,

    H. Liu, H. Yuan, Q. Liu, J. Hou, and J. Liu, “A Comprehensive Study and Comparison of Core Technologies for MPEG 3-D Point Cloud Compression,”IEEE Transactions on Broadcasting, vol. 66, no. 3, pp. 701-717, Dec. 2020

    work page 2020

  24. [24]

    3D Point Cloud Geometry Compression on Deep Learning,

    T. Huang and Y . Liu, “3D Point Cloud Geometry Compression on Deep Learning,” inProc. of the ACM International Conference on Multimedia, Nice, France, 2019, pp. 890–898

    work page 2019

  25. [25]

    Learning Convolutional Transforms for Lossy Point Cloud Geometry Compression,

    M. Quach, G. Valenzise, and F. Dufaux, “Learning Convolutional Transforms for Lossy Point Cloud Geometry Compression,” inProc. of the IEEE International Conference on Image Processing, Taipei, China, 2019, pp. 4320-4324

    work page 2019

  26. [26]

    Lossy Point Cloud Geometry Compression via Region-Wise Processing,

    W. Zhu, Y . Xu, D. Ding, Z. Ma, and M. Nilsson, “Lossy Point Cloud Geometry Compression via Region-Wise Processing,”IEEE Transactions on Circuits and Systems for Video Technology, vol. 31, no. 12, pp. 4575- 4589, Aug. 2021

    work page 2021

  27. [27]

    Multiscale Point Cloud Geometry Compression,

    J. Wang, D. Ding, Z. Li, and Z. Ma, “Multiscale Point Cloud Geometry Compression,” inProc. of the Data Compression Conference, Online, 2021, pp. 73-82

    work page 2021

  28. [28]

    Lossy Point Cloud Geometry Compression via End-to-End Learning,

    J. Wang, H. Zhu, H. Liu, and Z. Ma, “Lossy Point Cloud Geometry Compression via End-to-End Learning,”IEEE Transactions on Circuits and Systems for Video Technology, vol. 31, no. 12, pp. 4909-4923, Jan. 2021

    work page 2021

  29. [29]

    Point Cloud Geometry Compression Via Neural Graph Sampling,

    L. Gao, T. Fan, J. Wan, Y . Xu, J. Sun, and Z. Ma, “Point Cloud Geometry Compression Via Neural Graph Sampling,” inProc. of the IEEE International Conference on Image Processing, Anchorage, AK, USA, 2021, pp. 3373-3377

    work page 2021

  30. [30]

    Geometric Prior Based Deep Human Point Cloud Geometry Compression,

    X. Wu, P. Zhang, M. Wang, P. Chen, S. Wang and S. Kwong, “Geometric Prior Based Deep Human Point Cloud Geometry Compression,”IEEE Transactions on Circuits and Systems for Video Technology, vol. 34, no. 9, pp. 8794-8807, Sept. 2024

    work page 2024

  31. [31]

    Inter-Frame Compression for Dynamic Point Cloud Geometry Coding,

    A. Akhtar, Z. Li and G. Van der Auwera, “Inter-Frame Compression for Dynamic Point Cloud Geometry Coding,”IEEE Transactions on Image Processing, vol. 33, pp. 584-594, Jan. 2024

    work page 2024

  32. [32]

    Learning- Based Lossless Compression of 3D Point Cloud Geometry,

    D. T. Nguyen, M. Quach, G. Valenzise, and P. Duhamel, “Learning- Based Lossless Compression of 3D Point Cloud Geometry,” inProc. of the IEEE International Conference on Acoustics, Speech and Signal Processingn, Toronto, Canada, 2021, pp. 4220-4224

    work page 2021

  33. [33]

    Lossless Point Cloud Geometry and Attribute Compression Using a Learned Conditional Probability Model,

    D. T. Nguyen and A. Kaup, “Lossless Point Cloud Geometry and Attribute Compression Using a Learned Conditional Probability Model,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 33, no. 8, pp. 4337-4348, Jan. 2023

    work page 2023

  34. [34]

    Multiscale deep context modeling for lossless point cloud geometry compression,

    D. T. Nguyen, M. Quach, G. Valenzise, and P. Duhamel, “Multiscale deep context modeling for lossless point cloud geometry compression,” inProc. of the IEEE International Conference on Multimedia and Expo Workshops, Shenzhen, China, 2021, pp. 1-6

    work page 2021

  35. [35]

    Multiscale Latent-Guided Entropy Model for LiDAR Point Cloud Compression,

    T. Fan, L. Gao, Y . Xu, D. Wang, and Z. Li, “Multiscale Latent-Guided Entropy Model for LiDAR Point Cloud Compression,”IEEE Transactions on Circuits and Systems for Video Technology, vol. 33, no. 12, pp. 7857- 7869, May. 2023

    work page 2023

  36. [36]

    Sparse Tensor- Based Multiscale Representation for Point Cloud Geometry Compres- sion,

    J. Wang, D. Ding, Z. Li, X. Feng, C. Cao, and Z. Ma, “Sparse Tensor- Based Multiscale Representation for Point Cloud Geometry Compres- sion,”IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 45, no. 7, pp. 9055-9071, Jul. 2023

    work page 2023

  37. [37]

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

    J. Wang, R. Xue, J. Li, D. Ding, Y . Lin and Z. Ma, “A Versatile Point Cloud Compressor Using Universal Multiscale Conditional Coding – Part I: Geometry,”IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 47, no. 1, pp. 269-287, Jan. 2025

    work page 2025

  38. [38]

    K. You, T. Chen, D. Ding, M. S. Asif and Z. Ma, ”RENO: Real-Time Neural Compression for 3D LiDAR Point Clouds,” inProc. of IEEE/CVF Conference on Computer Vision and Pattern Recognition, Nashville, TN, USA, 2025, pp. 22172-22181

    work page 2025

  39. [39]

    OctSqueeze: Octree-Structured Entropy Model for LiDAR Compression,

    L. Huang, S. Wang, K. Wong, J. Liu, and R. Urtasun, “OctSqueeze: Octree-Structured Entropy Model for LiDAR Compression,” inProc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Seattle, Washington, USA, 2020, pp. 1313-1323

    work page 2020

  40. [40]

    Enhancing Context Models for Point Cloud Geometry Compression with Context Feature Residuals and Multi-Loss,

    C. Sun, H. Yuan, S. Li, X. Lu and R. Hamzaoui, “Enhancing Context Models for Point Cloud Geometry Compression with Context Feature Residuals and Multi-Loss,”IEEE Journal on Emerging and Selected Topics in Circuits and Systems, vol. 14, no. 2, pp. 224-234, Jun. 2024

    work page 2024

  41. [41]

    Enhancing Octree- Based Context Models for Point Cloud Geometry Compression With Attention-Based Child Node Number Prediction,

    C. Sun, H. Yuan, X. Mao, X. Lu and R. Hamzaoui, “Enhancing Octree- Based Context Models for Point Cloud Geometry Compression With Attention-Based Child Node Number Prediction,”IEEE Signal Processing Letters, vol. 31, pp. 1835-1839, Jul. 2024

    work page 2024

  42. [42]

    MuSCLE: Multi Sweep Compression of LiDAR using Deep Entropy Models,

    S. Biswas, J. Liu, K. Wong, S. Wang, and R. Urtasun, “MuSCLE: Multi Sweep Compression of LiDAR using Deep Entropy Models,” inProc. of the Advances in Neural Information Processing Systems, Palermo, Sicily, Italy, 2020, pp. 22170-22181

    work page 2020

  43. [43]

    R. Song, C. Fu, S. Liu and G. Li, ”Large-Scale Spatio-Temporal Attention Based Entropy Model for Point Cloud Compression,” inProc. of IEEE International Conference on Multimedia and Expo, Brisbane, Australia, 2023, pp. 2003-2008

    work page 2023

  44. [44]

    MDLPCC: Misalignment-aware dynamic LiDAR point cloud compression,

    A. Luo, L. Song, K. Nonaka, J. Liu, K. Unno, K. Matsuzaki, H. Sun, and J. Katto, “MDLPCC: Misalignment-aware dynamic LiDAR point cloud compression,”J. Vis. Commun. Image Represent., vol. 110, p. 104481, 2025

    work page 2025

  45. [45]

    X. Wang, K. Xu, X. Liu, J. Wan, Y . Guo and H. Wang, ”OctGLP-Net: Learning Octree-Structured Context Entropy Model With Global–Local Perception for Point Cloud Geometry Compression,”IEEE Transactions on Intelligent Transportation Systems, Early Access

    work page

  46. [46]

    X. Wang, Y . Zhang, T. Liu, X. Liu, K. Xu, J. Wan, ”Top- Net: Transformer-Efficient Occupancy Prediction Network for Octree- Structured Point Cloud Geometry Compression,” inProc. of IEEE/CVF Conference on Computer Vision and Pattern Recognition, Nashville, TN, USA, 2025, pp. 27305-27314

    work page 2025

  47. [47]

    Wang et al., ”GCFI-Net: Global-Local Cross-Spatial-Channel Feature Interaction Network for Point Cloud Geometry Compression,”IEEE Transactions on Mobile Computing, vol

    X. Wang et al., ”GCFI-Net: Global-Local Cross-Spatial-Channel Feature Interaction Network for Point Cloud Geometry Compression,”IEEE Transactions on Mobile Computing, vol. 24, no. 12, pp. 13663-13677, Dec. 2025

    work page 2025

  48. [48]

    X. Wang, K. Xu, B. Deng, Y . Guo and H. Wang, ”ASRL: Adaptive Sparse Representation Learning for LiDAR Point Cloud Geometry Com- pression,”IEEE Signal Processing Letters, vol. 32, pp. 3884-3888, 2025

    work page 2025

  49. [49]

    Cui, Z, Yuyang, F

    M. Cui, Z, Yuyang, F. Mingjian, L. Junhua, L. Yehua Ling, X Jiahao, ”GAEM: Graph-driven Attention-based Entropy Model for LiDAR Point Cloud Compression,”IEEE Transactions on Circuits and Systems for Video Technology, early access

    work page

  50. [50]

    The multivariate skew-normal distribution,

    A. AZZALINI, A. DALLA V ALLE, “The multivariate skew-normal distribution,”Biometrika, vol 83, pp 715–726, Dec, 1996

    work page 1996

  51. [51]

    Accurate and robust registration of low overlapping point clouds,

    J. Yang, M. Zhao, Y . Wu, X Jia, “Accurate and robust registration of low overlapping point clouds,”Computers & Graphics, 2024

    work page 2024

  52. [52]

    Long Short-Term Memory,

    S. Hochreiter and J. Schmidhuber, “Long Short-Term Memory,”Neural Computation, vol. 9, no. 8, pp. 1735-1780, Nov. 1997

    work page 1997

  53. [53]

    Model-Based Encoding Parameter Optimization for 3D Point Cloud Compression,

    Q. Liu, H. Yuan, J. Hou, H. Liu and R. Hamzaoui, “Model-Based Encoding Parameter Optimization for 3D Point Cloud Compression,” in Proc. of the Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, Honolulu, HI, USA, 2018, pp. 1981- 1986

    work page 2018

  54. [54]

    Frame-Level Rate Control for Geometry- Based LiDAR Point Cloud Compression,

    L. Li, Z. Li, S. Liu and H. Li, “Frame-Level Rate Control for Geometry- Based LiDAR Point Cloud Compression,”IEEE Transactions on Multi- media, vol. 25, pp. 3855-3867, Apr, 2023

    work page 2023

  55. [55]

    Rate Control for Geometry-Based LiDAR Point Cloud Compression via Multi- Factor Modeling,

    L. Hou, L. Gao, Q. Zhang, Y . Xu, J. N. Hwang and D. Wang, “Rate Control for Geometry-Based LiDAR Point Cloud Compression via Multi- Factor Modeling,”IEEE Transactions on Broadcasting, pp. 1-13, Oct. 2024

    work page 2024

  56. [56]

    SemanticKITTI: A Dataset for Semantic Scene Under- standing of LiDAR Sequences,

    J. Behley, M. Garbade, A. Milioto, J. Quenzel, S. Behnke, C. Stachniss, and J. Gall, “SemanticKITTI: A Dataset for Semantic Scene Under- standing of LiDAR Sequences,” inProc. of the IEEE/CVF International Conference on Computer Vision, Seoul, Korea, 2019, pp. 9297-9307

    work page 2019

  57. [57]

    Fordcampus vision and lidar data set,

    G. Pandey, J. R. McBride, and R. M. Eustice, “Fordcampus vision and lidar data set,”The International Journal of Robotics Research, vol. 30, no. 13, pp. 1543–1552, Mar. 2011

    work page 2011

  58. [58]

    Common test con- ditions for point cloud compression,

    MPEG 3D Graphics Coding and Haptics Coding, “Common test con- ditions for point cloud compression,”ISO/IECJTC1/SC29/WG11 MPEG input document N17229, Macau, Jan. 2023

    work page 2023

  59. [59]

    Updates and Integra- tion of Evaluation Metric Software for PCC,

    MPEG 3D Graphics Coding and Haptics Coding, “Updates and Integra- tion of Evaluation Metric Software for PCC,”ISO/IECJTC1/SC29/WG11 MPEG input document M40522, Macau, Apr. 2017

    work page 2017

  60. [60]

    Enhanced G-PCC per- formance evaluation and anchor results,

    MPEG Coding for 3D Graphics and Haptics, “Enhanced G-PCC per- formance evaluation and anchor results,” inISO/IEC JTC1/SC29/WG07 MPEG output document N01233, Online, Jun. 2025

    work page 2025

  61. [61]

    V oxel R-CNN: Towards High Performance V oxel-based 3D Object Detection

    J. Deng, S. Shi, P. Li, W. Zhou, Y . Zhang, and H. Li, “V oxel R-CNN: Towards High Performance V oxel-based 3D Object Detection”, inProc. of the AAAI Conference on Artificial Intelligence, online, 2021, pp. 3954- 3962

    work page 2021