Research map / Point clouds / Point cloud classification
OctAttention: Octree-Based Large-Scale Contexts Model for Point Cloud Compression
Chunyang Fu, Ge Li, Rui Song, Wei Gao, Shan Liu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use OctAttention instead of G-PCC | Object point cloud compression · MVUB average | bpp: 0.95 -> 0.76 | Improved |
| Use OctAttention instead of VoxelDNN | Object point cloud compression · MVUB average | bpp: 0.81 -> 0.76 | Improved |
| Use OctAttention instead of G-PCC | Object point cloud compression · MPEG 8i average | bpp: 0.99 -> 0.64 | Improved |
| Use OctAttention instead of VoxelDNN | Object point cloud compression · MPEG 8i average | bpp: 0.73 -> 0.64 | Improved |
| Increase context window from N=8 to N=1024 (ablation) | OctAttention · SemanticKITTI, D=10 | bpp: 1.086 -> 0.939 | Improved |
| Increase context window from N=512 to N=1024 (ablation) | OctAttention · SemanticKITTI, D=10 | encode time / K nodes: 0.0028 -> 0.0015 | Hurt |
| Add sibling context to an MLP backbone (ablation) | OctAttention · SemanticKITTI, D=10 | bpp: 1.315 -> 1.117 | Improved |
| Add sibling context to an attention backbone (ablation) | OctAttention · SemanticKITTI, D=10 | bpp: 1.289 -> 0.959 | Improved |
| Use attention instead of MLP with sibling context (ablation) | OctAttention · SemanticKITTI, D=10 | bpp: 1.117 -> 0.959 | Improved |
| Use attention instead of MLP without sibling context (ablation) | OctAttention · SemanticKITTI, D=10 | bpp: 1.315 -> 1.289 | Mixed |
| Use the mask operation to encode multiple nodes in parallel (ablation) | OctAttention · Octree node coding | encoding time: coding time reduces by N_{0} times | Improved |
| Use OctAttention instead of VoxelDNN | Object point cloud compression · MPEG 8i average encoding time | encoding time: 2459 -> 0.41 | Improved |
Later papers that compared with this method
7 later accepted papers report OctAttention as a baseline in their results tables; 4 of them report a better number on most of the results they share with it.
- R²D-LPCC: Relevance-Ranking Guided Region-Adaptive Dynamic LiDAR Point Cloud Compression (AAAI 2026 · better on 1 of 2 shared results)
- PACE: Post-Causal Entropy Modeling for Learned LiDAR Point Cloud Compression (ICML 2026 · better on 1 of 1 shared result)
- Perceive More with Less: LiDAR Point Cloud Compression at Just Recognizable Distortion for 3D Scene Understanding (AAAI 2026 · better on 1 of 1 shared result)
- UniPCGC: Towards Practical Point Cloud Geometry Compression via an Efficient Unified Approach (AAAI 2025 · better on 1 of 1 shared result)
- Efficient Hierarchical Entropy Model for Learned Point Cloud Compression (CVPR 2023 · better on 1 of 5 shared results)
- OctFormer: Efficient Octree-Based Transformer for Point Cloud Compression with Local Enhancement (AAAI 2023 · better on 1 of 3 shared results)
- Sparse Tensor-Based Multiscale Representation for Point Cloud Geometry Compression (TPAMI 2022 · better on 2 of 2 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Point Cloud Compression with Sibling Context and Surface Priors (ECCV 2022 · similarity 0.93)
- VoxelContext-Net: An Octree Based Framework for Point Cloud Compression (CVPR 2021 · similarity 0.92)
- OctSqueeze: Octree-Structured Entropy Model for LiDAR Compression (CVPR 2020 · similarity 0.91)
- OctFormer: Efficient Octree-Based Transformer for Point Cloud Compression with Local Enhancement (AAAI 2023 · similarity 0.91)
- TopNet: Transformer-Efficient Occupancy Prediction Network for Octree-Structured Point Cloud Geometry Compression (CVPR 2025 · similarity 0.88)
- Multi-Scale Neighborhood Occupancy Masked Autoencoder for Self-Supervised Learning in LiDAR Point Clouds (CVPR 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
