AI Paper Atlas

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

From Scene to Object: Enhancing Open-Vocabulary Object Detection via Foreground-Background Context Reasoning

Yanqi Li, Jianwei Niu, Ningbo Gu, Tao Ren

AAAI 2026 · Detection and segmentation › Open-set and few-shot detection · approach: Open-vocabulary object detection · 1 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 BFDet pseudo-labelsVLDet · OV-COCOAP novel: 32.0 -> 35.1Improved
add BFDet pseudo-labelsViLD, Detic, RegionCLIP, BARON · OV-COCOAP novel: 27.6 -> 29.9; 27.8 -> 30.3; 26.8 -> 29.5; 35.1 -> 37.8Improved
add BFDet pseudo-labelsfive detectors · OV-COCOAP known: 59.5 -> 62.1 (ViLD)Improved
add BFDet pseudo-labelsViLD, RegionCLIP, VLDet, BARON · OV-LVIS (RN50)mask AP novel: 16.6 -> 19.1; 17.1 -> 19.5; 21.7 -> 23.6; 19.2 -> 20.7Improved
add BFDet pseudo-labelsVLDet, Detic · OV-LVIS (Swin-B)mask AP novel: 26.3 -> 28.1; 23.9 -> 25.4Improved
add Scene-Aware Background Reasoning vs Foreground Objects Alignment (ablation)VLDet · OV-COCOAP novelImproved
mask out low-confidence objects when embedding background (ablation)BFDet (VLDet) · OV-COCOAP novel: novel -0.3, known +0.5 without maskMixed
use incorrect background in object prompt (ablation)BFDet (VLDet) · OV-COCOAP novel / AP all: -3.7 AP novel, -3.8 AP allHurt
tune thresholds alpha=0.6 and beta=0.5 (ablation)BFDet (VLDet) · OV-COCOAPImproved
BFDet-enhanced BARON vs best prior methodBARON-BFDet · OV-COCOAP novel: 37.8 vs CAKE 39.1No clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 0 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