AI Paper Atlas

Research map / Point clouds / Segmentation and detection

3D-MPA: Multi-Proposal Aggregation for 3D Semantic Instance Segmentation

CVPR 2020 · Point clouds › Segmentation and detection · approach: Point cloud semantic segmentation · 237 citations (OpenAlex)

Check your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's title only (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
Replace NMS with aggregation of multiple proposals by position (ablation)3D-MPA · ScanNetV2 valmAP: 47.5 -> 52.4; +4.9 mAPHelped
Use learned embedding features rather than proposal positions for aggregation (ablation)3D-MPA · ScanNetV2 valmAP: 52.4 -> 56.7Helped
Use learned geometric features rather than proposal positions for aggregation (ablation)3D-MPA · ScanNetV2 valmAP: 52.4 -> 57.8; +5.4 mAPHelped
Use geometric rather than embedding aggregation features (ablation)3D-MPA · ScanNetV2 valmAP: 56.7 -> 57.8; +1.1 mAPHelped
Add a 10-layer GCN for proposal consolidation (ablation)3D-MPA with geometric aggregation features · ScanNetV2 valmAP: 57.8 -> 59.1; +1.3 mAPHelped
Replace NMS with multi-proposal aggregation and add the GCN (ablation)3D-MPA · ScanNetV2 valmAP: 47.5 -> 59.1; +11.6 mAPHelped
Sample proposal positions randomly instead of using Farthest Point Sampling3D-MPA · 3D point-cloud proposal generationproposal qualityHelped

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