AI Paper Atlas

Research map / Point clouds / Segmentation and detection

Clusterformer: Cluster-based Transformer for 3D Object Detection in Point Clouds

Yu Pei, Xian Zhao, Hao Li, Jingyuan Ma, Jingwei Zhang, Shiliang Pu

ICCV 2023 · Point clouds › Segmentation and detection · approach: Single-stage point-voxel detection · 15 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
replace cluster feature pooling with intra-cluster cross-attention (ablation)Clusterformer · Waymo val (ablation)Overall L2 mAPH: 68.6 -> 70.0Improved
add cluster-wise self-attention between queries (ablation)Clusterformer · Waymo val (ablation)Overall L2 mAPH: 70.0 -> 70.4Improved
add Query2Key strategy (ablation)Clusterformer · Waymo val (ablation)Overall L2 mAPH: 70.4 -> 71.2Improved
initialize queries from cluster centers instead of farthest point sampling (ablation)Clusterformer · Waymo val (ablation), vehicleL2 mAPH: 61.7 -> 65.3Improved
limit interaction to the same cluster rather than distance-wise range (ablation)Clusterformer · Waymo val (ablation), vehicleL2 mAPH: 65.3 -> 66.9Improved
global-wise interaction range for queries (ablation)Clusterformer · Waymo val (ablation), vehicleL2 mAPH: 66.9 -> 49.4 (cluster center init)Hurt
zero-initialized queries with intra-cluster interaction (ablation)Clusterformer · Waymo val (ablation), vehicleL2 mAPH: 66.9 -> 65.9Mixed
increase decoder layers with Query2Key (ablation)Clusterformer · Waymo val (ablation)Overall L2 mAPH: +1.6 with Query2Key vs +0.8 without (2 to 4 layers)Improved
remove the local maximum operation in cluster generation (ablation)Clusterformer · Waymo val (ablation)Overall L2 mAPH: 71.2 -> 69.9Hurt
remove NMS with Max-IoU matching (ablation)Clusterformer · Waymo val (ablation)Overall L2 mAPH: 71.2 -> 68.1Hurt
use Hungarian matching so NMS is not needed (ablation)Clusterformer · Waymo val (ablation)Overall L2 mAPH: 68.1 -> 70.2 (no NMS)Improved
Clusterformer versus Centerformer in latency and memoryClusterformer · Waymo vallatency / inference memory: 123ms -> 103ms; 10.4G -> 5.7GImproved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most 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

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