AI Paper Atlas

Research map / 3D vision / 3D generation

SpaCeFormer: Fast Proposal-Free Open-Vocabulary 3D Instance Segmentation

Chris Choy, Junha Lee, Chunghyun Park, Minsu Cho, Jan Kautz

ICML 2026 · 3D vision › 3D generation · approach: Learning 3D reconstruction from images · 0 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 Morton curve attention to window attention (ablation)SpaCeFormer backbone · ScanNet200 valmAP: 0.09517 -> 0.11092Improved
add window attention to Morton-only attention (ablation)SpaCeFormer backbone · ScanNet200 valmAP: 0.09470 -> 0.11092Improved
add Morton attention to window-only on class-agnostic AP (ablation)SpaCeFormer backbone · ScanNet200 valAP: 0.25466 -> 0.25178No clear effect
use RoPE in decoder instead of no positional encoding (ablation)SpaCeFormer decoder · ScanNet200 valmAP: 5.97 -> 7.60Improved
use RoPE instead of absolute positional embedding (ablation)SpaCeFormer decoder · ScanNet200 valmAP: 5.95 -> 7.60Improved
use RoPE instead of learnable positional bias (ablation)SpaCeFormer decoder · ScanNet200 valmAP: 6.46 -> 7.60Improved
use learned queries instead of random init (ablation)SpaCeFormer decoder · ScanNet200 valclass-agnostic AP: 0.14635 -> 0.19819Improved
use learned queries instead of FPS init (ablation)SpaCeFormer decoder · ScanNet200 valclass-agnostic AP: 0.17970 -> 0.19819Improved
use learned queries instead of FPS init on class-wise mAP (ablation)SpaCeFormer decoder · ScanNet200 valmAP: 0.07164 -> 0.06443Hurt
multi-view mask clustering instead of naive SAM2 2D-to-3D liftingSpaCeFormer-3M masks · 312 ScanNet train scenesrecall@0.5: 2.5% -> 54.3%Improved
replace Mask3D proposals with Segment3D open-vocab proposalsMosaic3D · ScanNet200mAP: 11.8 -> 2.7Hurt
replace Segment3D proposals with proposal-free decoderMosaic3D / SpaCeFormer · ScanNet200mAP: 2.7 -> 11.1Improved

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