.png){kind=icon}
Return
RegionDiffusion: Generative data augmentation for object detection with diffusion models
.png){kind=icon}
PRE
AI
DOI：10.1016/j.neucom.2025.131115.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• We propose RegionDiffusion, a novel layout-to-image model with outstanding generation and zero-shot capabilities enabled by region self-attention. • We introduce a comprehensive generative data augmentation pipeline utilizing a series of preprocessing and postprocessing steps, which significantly enhance the performance of object detectors across general and specific domains.
Keywords:
RegionDiffusion
layout-to-image model
region self-attention
generative data augmentation
object detection
Journal
IF:
6.5
Papers:
2.5W
Citations:
6.5W
Organization
No organization information available
Cited Papers
No cited papers available