AI Paper Atlas

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

SegPrompt: Boosting Open-World Segmentation via Category-Level Prompt Learning

Muzhi Zhu, Hengtao Li, Hao Chen, Chengxiang Fan, Weian Mao, Chenchen Jing, Yifan Liu, Chunhua Shen

ICCV 2023 · Detection and segmentation › Open-set and few-shot detection · approach: Amodal and open-world instance segmentation · 21 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
class-agnostic training instead of class-aware (ablation)Mask2Former · LVIS-OWAR_all: 40.8 -> 42.5Improved
class-agnostic training on known categories (ablation)Mask2Former · LVIS-OWAR_known: 50.2 -> 49.9No clear effect
add SegPrompt category-level prompt auxiliary supervisionclass-agnostic Mask2Former · LVIS-OWAR_all: 42.5 -> 44.9Improved
add SegPrompt auxiliary supervisionclass-agnostic Mask2Former · LVIS-OWAR_unseen: 47.4 -> 50.3Improved
use prompt-based per-category supervision instead of vanilla Mask2Former auxiliary classification (ablation)SegPrompt · LVIS-OWAR_all: 43.0 -> 44.1Improved
add example supervision (ablation)SegPrompt · LVIS-OWAR_all: 44.1 -> 44.9Improved
add example supervision (ablation)SegPrompt · COCO to LVISAP_l: 28.0 -> 24.5Mixed
add class-specific embedding s to candidate queries (ablation)SegPrompt · LVIS-OWAR_all: 43.3 -> 44.1Improved
add prompt p to candidate queries (ablation)SegPrompt · LVIS-OWAR_all: 42.8 -> 44.1Improved
use reference attention instead of shared global self-attention (ablation)SegPrompt · LVIS-OWAR_all: 35.7 -> 44.1Improved
make candidate queries of different classes invisible to each other (ablation)SegPrompt RefAttn · LVIS-OWAR_all: 44.0 -> 44.1No clear effect
apply SegPrompt in full LVIS supervised trainingMask2Former · LVIS valAP@100: 20.1 -> 22.8Improved

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