AI Paper Atlas

Research map / Point clouds / Segmentation and detection

Hierarchical Aggregation for 3D Instance Segmentation

Shaoyu Chen, Jiemin Fang, Qian Zhang, Wenyu Liu, Xinggang Wang

ICCV 2021 · Point clouds › Segmentation and detection · approach: Scene completion and instance segmentation · 188 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
add intra-instance prediction network (ablation)HAIS · ScanNet v2 valAP50: 61.0 -> 63.4Improved
add intra-instance prediction network (ablation)HAIS · ScanNet v2 valAP: 39.8 -> 42.5Improved
add intra-instance prediction network (ablation)HAIS · ScanNet v2 valAP25: 74.6 -> 74.9Improved
add set aggregation (ablation)HAIS · ScanNet v2 valAP50: 63.4 -> 64.1Improved
add set aggregation (ablation)HAIS · ScanNet v2 valAP: 42.5 -> 43.5Improved
add set aggregation (ablation)HAIS · ScanNet v2 valAP25: 74.9 -> 75.6Improved
use masks to filter background points and compute IoU supervision for scoring (ablation)HAIS · ScanNet v2 valAP: 41.9 -> 43.5Improved
use masks to filter background points and compute IoU supervision for scoring (ablation)HAIS · ScanNet v2 valAP50: 63.1 -> 64.1Improved
filter mask training samples to IoU > 0.5 instead of using all (ablation)HAIS · ScanNet v2 valAP: 42.8 -> 43.5Improved
filter mask training samples to IoU > 0.5 instead of using all (ablation)HAIS · ScanNet v2 valAP25: 74.4 -> 75.6Improved
use HAIS instead of previous SOTA on test setHAIS · ScanNet v2 testAP50: 63.8 -> 69.9Improved
use HAIS instead of PointGroupHAIS · ScanNet v2 valper-frame inference time (ms): 452 -> 410Improved

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