.png){kind=icon}
.png){kind=icon}
Return
GPS-SAM: text-driven Grounded Polyp Segmentation SAM
.png){kind=icon}
PRE
AI
DOI：10.1016/j.neucom.2026.134063.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Learnable text prompts boost polyp segmentation accuracy in colonoscopy images • GPS-SAM improves generalization for unseen polyps in clinical scenarios • Language-guided annotation enhances precision of SAM in colorectal image analysis • Fusion of visual and textual cues enables robust polyp detection across datasets • New method outperforms static prompts by capturing complex polyp characteristics
Keywords:
polyp segmentation
text prompts
SAM
generalization
language-guided annotation