AI Paper Atlas

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

AAAI 2022 · Point clouds › Point cloud classification · approach: Scene flow and dynamic point clouds · 196 citations (OpenAlex)

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 changedWhereMeasuredEffect
Use OctAttention instead of G-PCCObject point cloud compression · MVUB averagebpp: 0.95 -> 0.76Improved
Use OctAttention instead of VoxelDNNObject point cloud compression · MVUB averagebpp: 0.81 -> 0.76Improved
Use OctAttention instead of G-PCCObject point cloud compression · MPEG 8i averagebpp: 0.99 -> 0.64Improved
Use OctAttention instead of VoxelDNNObject point cloud compression · MPEG 8i averagebpp: 0.73 -> 0.64Improved
Increase context window from N=8 to N=1024 (ablation)OctAttention · SemanticKITTI, D=10bpp: 1.086 -> 0.939Improved
Increase context window from N=512 to N=1024 (ablation)OctAttention · SemanticKITTI, D=10encode time / K nodes: 0.0028 -> 0.0015Hurt
Add sibling context to an MLP backbone (ablation)OctAttention · SemanticKITTI, D=10bpp: 1.315 -> 1.117Improved
Add sibling context to an attention backbone (ablation)OctAttention · SemanticKITTI, D=10bpp: 1.289 -> 0.959Improved
Use attention instead of MLP with sibling context (ablation)OctAttention · SemanticKITTI, D=10bpp: 1.117 -> 0.959Improved
Use attention instead of MLP without sibling context (ablation)OctAttention · SemanticKITTI, D=10bpp: 1.315 -> 1.289Mixed
Use the mask operation to encode multiple nodes in parallel (ablation)OctAttention · Octree node codingencoding time: coding time reduces by N_{0} timesImproved
Use OctAttention instead of VoxelDNNObject point cloud compression · MPEG 8i average encoding timeencoding time: 2459 -> 0.41Improved

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.

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

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy