Safety
Enhancing industrial safety AI with synthetic data on Amazon SageMaker AI
Learn how to build a synthetic data augmentation pipeline on Amazon SageMaker AI and Amazon Rekognition that generates photo-realistic, auto-labeled training images for industrial safety AI. This appr
Learn how to build a synthetic data augmentation pipeline on Amazon SageMaker AI and Amazon Rekognition that generates photo-realistic, auto-labeled training images for industrial safety AI. This approach improved person detection by up to 160% without manual annotation or hazardous data collection near heavy machinery.
Related
- ScoreMix: Synthetic Data Generation by Score Composition in Diffusion Models Improves Recognition
- A Data Efficiency Study of Synthetic Fog for Object Detection Using the Clear2Fog Pipeline
- Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting
Source: AWS ML Blog | 2026-09-17