Presentation + Paper
7 June 2024 Integrating synthetic data validation and quality benchmarks into a continuous integration/continuous delivery (CI/CD) data-generation pipeline
Eric Fiterman, Kenneth Brown, Daniel Mallia, Justin Tornetta
Author Affiliations +
Abstract
Measuring and validating the quality, visual fidelity, and performance of synthetic image data is an advanced and evolving subject. This study will present an overview of various methods and approaches for measuring the visual quality and fidelity of synthetic image data. Cignal will present an overview of approaches, including established industry standards, such as ASTM E1695 and ANSI N42.45, statistical methods, and neural-network based approaches. Lastly, Cignal will discuss how these approaches may be integrated into a Continuous Integration/ Continuous Delivery (CI/CD) data generation pipeline to monitor and improve data quality and predicted performance for Artificial Intelligence/Machine Learning (AI/ML) use cases.
Conference Presentation
(2024) Published by SPIE. Downloading of the abstract is permitted for personal use only.
Eric Fiterman, Kenneth Brown, Daniel Mallia, and Justin Tornetta "Integrating synthetic data validation and quality benchmarks into a continuous integration/continuous delivery (CI/CD) data-generation pipeline", Proc. SPIE 13043, Anomaly Detection and Imaging with X-Rays (ADIX) IX, 130430C (7 June 2024); https://doi.org/10.1117/12.3014548
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Image quality

Data modeling

X-rays

X-ray imaging

Education and training

Visualization

Machine learning

Back to Top