AI Paper Atlas

Research map / 3D vision / Gaussian splatting

Chorus: Multi-Teacher Pretraining for Holistic 3D Gaussian Scene Encoding

Yue Li, Qi Ma, Runyi Yang, Mengjiao Ma, Bin Ren, Nikola Popovic, Nicu Sebe, Theo Gevers, Luc Van Gool, Danda Pani Paudel, Martin R. Oswald

CVPR 2026 · 3D vision › Gaussian splatting · approach: Compact and efficient Gaussian splatting · 5 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 DINO teacher to the language teacher (ablation)Chorus (3DGS) · ScanNet++ v2-trained, zero-shot seg on ScanNet++ valf-mIoU: 27.1 -> 29.4Improved
add PE-Spatial teacher on top of Lang+DINO (ablation)Chorus (3DGS) · InteriorGS zero-shot segf-mIoU: 9.3 -> 11.4Improved
add PE-Spatial teacher on top of Lang+DINO (ablation)Chorus (3DGS) · ScanNet++ val zero-shot segf-mIoU: 29.4 -> 29.6No clear effect
introduce PE-Spatial only in second half of training instead of from the start (ablation)Chorus pretraining · ScanNet++ val zero-shot segsegmentation scoreImproved
use SmoothL1 loss, 3DGS-aware augmentations and instance contrastive term (ablation)Chorus pretraining · ScanNet++ val zero-shot segsegmentation scoreImproved
replace SceneSplat with Chorus 3D backbone in GaussianVLMGaussianVLM · Nr3DSim: 20.8 -> 22.5Improved
replace SceneSplat with Chorus in GaussianVLMGaussianVLM · ScanQAMETEOR: 22.9 -> 22.5 (M); 14.4 -> 14.8 (EM1)Mixed
Chorus multi-teacher pretraining vs Sonata self-supervised, linear probingpoint-cloud encoder · ScanNet200 valmIoU: 28.8 -> 36.0Improved
Chorus vs Sonata in linear-probe instance segmentation on ScanNetpoint-cloud encoder · ScanNet val instance segmAP25: 72.6 -> 66.6Hurt
Chorus vs Sonata with decoder probing on ScanNet semantic segpoint-cloud encoder · ScanNet valmIoU: 77.3 -> 75.0Hurt
add render-and-distill adaptation on 100 extra InteriorGS scenesChorus · InteriorGS linear probingmIoU: +2.7% mIoUImproved
Chorus pretraining as input-noise-robust featuresChorus point-cloud variant · ScanNet++ instance retrieval PC to noisy PCR@1: 79.8% -> 85.4%Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 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