AI Paper Atlas

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

SIGMA++: Improved Semantic-complete Graph Matching for Domain Adaptive Object Detection

Wuyang Li, Xinyu Liu, Yixuan Yuan

TPAMI 2023 · Detection and segmentation › Open-set and few-shot detection · approach: Domain-adaptive object detection · 66 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 Graph-embedded Semantic Completion module (ablation)GA baseline · Cityscapes to Foggy Cityscapes (VGG-16)mAP: 35.3 -> 41.8Improved
add Domain-guided Node Completion (ablation)GSC · Cityscapes to Foggy Cityscapes (VGG-16)mAP: 40.4 -> 41.8Improved
use graph-guided memory bank instead of common buffer (ablation)GSC · Cityscapes to Foggy Cityscapes (VGG-16)mAP: 41.0 -> 41.8Improved
add Node Discriminator (ablation)GSC · Cityscapes to Foggy Cityscapes (VGG-16)mAP: 39.4 -> 41.8Improved
add Bipartite Graph Matching adaptor (ablation)GSC · Cityscapes to Foggy Cityscapes (VGG-16)mAP: 41.8 -> 43.5Improved
add Cross Graph Interaction (ablation)BGM · Cityscapes to Foggy Cityscapes (VGG-16)mAP: 42.8 -> 43.5Improved
add Semantic-aware Node Affinity (ablation)BGM · Cityscapes to Foggy Cityscapes (VGG-16)mAP: 42.6 -> 43.5Improved
add Structure-aware Matching Loss (ablation)BGM · Cityscapes to Foggy Cityscapes (VGG-16)mAP: 42.2 -> 43.5Improved
increase sampled nodes from 20 to 200 per feature map (ablation)SIGMA · Cityscapes to Foggy Cityscapes (VGG-16)mAP: 39.0 -> 43.9Improved
increase sampled nodes from 200 to 500 (ablation)SIGMA · Cityscapes to Foggy Cityscapes (VGG-16)mAP: 43.9 -> 42.6Hurt
single matching instead of multiple matching (ablation)SIGMA · Cityscapes to Foggy Cityscapes (VGG-16)mAP: 43.1 -> 43.5Improved
add FS and QC terms to TE matching loss (ablation)SIGMA · Cityscapes to Foggy Cityscapes (VGG-16)mAP: 42.1 -> 43.5Improved

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