.png){kind=icon}
.png){kind=icon}
Return
A Scalable Pipeline Combining Procedural 3D Graphics and Guided Diffusion for Photorealistic Synthetic Training Data Generation in White Button Mushroom Segmentation
.png){kind=icon}
OA
AI
A
P
DOI：10.1016/j.compag.2026.112269.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Workflow generating annotated, photo-real, synthetic instance segmentation datasets. • 3D graphics and diffusion models for high realism without manual shader setup. • Full control of geometry and annotation while implicitly modeling visual properties. • Two synthetic Agaricus Bisporus mushroom datasets with instance masks and depth maps. • Synthetic data-trained models perform like state-of-the-art on real-world data.
Keywords:
Generative AI
Computer graphics
Diffusion models
Synthetic dataset
Mushroom segmentation
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
8.9
Papers:
9.9K
Citations:
4.8W
Organization
No organization information available
Cited Papers