AI Paper Atlas

Research map / Detection and segmentation / Open-set and few-shot detection

Synthetic Object Compositions for Scalable and Accurate Learning in Detection, Segmentation, and Grounding

Weikai Huang, Jieyu Zhang, Taoyang Jia, Chenhao Zheng, Ziqi Gao, Jae Sung Park, Ranjay Krishna

CVPR 2026 · Detection and segmentation › Open-set and few-shot detection · approach: Amodal and open-world instance segmentation · 4 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 50K SOC images to O365+GoldGMM-Grounding-DINO · LVIS APAP: 20.1 -> 29.8Improved
Add 20M GRIT vs 50K SOCMM-Grounding-DINO · LVIS APAP: +7.0 (GRIT) vs +9.7 (SOC-50K)Improved
Scale SOC from 100K to 400KMM-Grounding-DINO · OdinW-35avg AP: 21.0 -> 22.8Improved
Add 50K SOCMM-Grounding-DINO · OdinW-35avg AP: 20.3 -> 20.5No clear effect
Add 100K SOC to GRIT+V3DetMM-Grounding-DINO · LVIS APrareAPrare: 23.6 -> 29.8Improved
Add V3Det to O365+GoldGMM-Grounding-DINO · gRefCOCO no-target accuracyNacc: 89.3 -> 89.3No clear effect
Add SOC-FC-100KMM-Grounding-DINO · gRefCOCO no-target accuracyNacc: 89.3 -> 97.7Improved
Add 50K SOC to LVIS trainingAPE · LVIS v1 valAPrare: 40.87 -> 44.70Improved
Add 3D geometric layout vs random 2D layout (ablation)SOC pipeline · COCO instance segAP: 9.07 -> 10.03Improved
Add camera configuration augmentation (ablation)SOC pipeline · COCO instance segAP: 10.03 -> 10.58Improved
Add generative harmonization (inpainting+relighting) (ablation)SOC pipeline · COCO instance segAP: 6.28 -> 10.58Improved
Add mask-area-weighted blending (ablation)SOC pipeline · COCO instance segAP: 10.58 -> 12.79Improved

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