AI Paper Atlas

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

SegEarth-R2: Towards Comprehensive Language-guided Segmentation for Remote Sensing Images

Zepeng Xin, Kaiyu Li, Luodi Chen, Wanchen Li, Xiao Yuchen, Hui Qiao, Weizhan Zhang, Deyu Meng, Xiangyong Cao

CVPR 2026 · Detection and segmentation › Open-set and few-shot detection · approach: Remote sensing segmentation and change · 12 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 spatial attention supervision (lambda_S=0.01) (ablation)SegEarth-R2 · RRSIS-D testgIoU: 66.6 -> 67.9Improved
add spatial attention supervision (lambda_S=0.01) (ablation)SegEarth-R2 · RefSegRS testgIoU: 70.4 -> 71.8Improved
add spatial attention supervision (lambda_S=0.01) (ablation)SegEarth-R2 · EarthReason testgIoU: 72.9 -> 73.5Improved
raise lambda_S to 0.1 (ablation)SegEarth-R2 · EarthReason testgIoU: 73.5 -> 71.8Hurt
use a single [SEG] query instead of 100 queries with matching (ablation)SegEarth-R2 · RRSIS-D valgIoU: 66.4 -> 68.8Improved
use a single [SEG] query instead of 100 queries (ablation)SegEarth-R2 · inference times/image: 0.82 -> 0.54Improved
use a single [SEG] query instead of 100 queries (ablation)SegEarth-R2 · computeTFLOPs: 8.4 -> 6.1Improved
use Mask2Former decoder with Swin-B instead of SAM decoder with ViT-H (ablation)SegEarth-R2 · EarthReason testgIoU: 65.4 -> 73.5Improved
use Mask2Former decoder with Swin-B instead of SAM 2 decoder with ViT-H (ablation)SegEarth-R2 · RefSegRS testgIoU: 64.2 -> 71.8Improved
add spatial attention supervisionSegEarth-R2 · LaSeRS part-levelgIoUImproved
use 3B model against 8B GeoPixelSegEarth-R2-3B · LaSeRS multi-targetgIoU: 49.2 -> 38.3Hurt

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