EUREKA Xecs TASTI Application-tailored Synthetic Image Generation

Purpose and goal:

The main goal of TASTI in general is to develop a modular framework of transferable technology to innovate synthetic image generation and tailor it towards applications to enable synthetic data generation across healthcare, automotive industry, material production and agriculture domains.

The Swedish part of the consortium, consisting of Mentice AB, will support all work packages and lead one of them (WP3) to create the framework, however only apply it towards the Healthcare use case. Expected results and effects:

The TASTI project will open several new doors towards computer-vision applications by creating training data that was previously not possible to produce, not least in the healthcare use case where Mentice has its focus. After the project, healthcare in Sweden and Europe may be improved by increasing the quality of advanced endovascular procedures with planning tools and decision support tools for patient-specific care.

Today, there is far from sufficient training data for such advanced tools, which TASTI intends to solve. Approach and implementation: The project has been divided into a selection of work packages.

WP1 targets requirements, architecture and infrastructure. WP2 model and image preparation of data coming from the use case.

WP3 generates the synthetic images using physics or generative model, WP4 focuses on computing solutions for optimizations. WP5 Integrates solution to provide a demonstration, also evaluates solution against the requirements defined in WP1.

WP6 looks at how the solution can be utilised and WP7 Project management aim to maximize the project’s outcome and facilitate collaboration.