.png){kind=icon}
Return
Prior tokenization-based interactive segmentation with Vision Transformers
.png){kind=icon}
PRE
AI
DOI：10.1016/j.patcog.2025.112361.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Prior tokens handle semantic features in interactive information better than distance maps. • Under discriminative constraints, prior tokens distinguish target and background semantic divergence more easily. • Cross-attention tightly aligns image patch tokens with user intent. • The register method effectively addresses artifacts in interactive segmentation.
Keywords:
prior tokens
cross-attention
interactive segmentation
semantic features
discriminative constraints
Journal
IF:
7.6
Papers:
1.3W
Citations:
4.5W
Organization
Cited Papers
A fully convolutional two-stream fusion network for interactive image segmentation.png)
NEURAL NETWORKS
OAAI
NEURAL NETWORKSIF6.3
OAAIShareSaveShareSaveShareSaveText-video retrieval re-ranking via multi-grained cross attention and frozen image encodersPATTERN RECOGNITION
PATTERN RECOGNITIONIF7.6
PREAI
ShareSaveShareSaveShareSaveVolumetric memory network for interactive medical image segmentationMEDICAL IMAGE ANALYSIS
OAAI
MEDICAL IMAGE ANALYSISIF11.8
OAAIShareSaveShareSave
View more