What would it cost to license or deploy this technology?+
The project was publicly funded with EUR 3,937,850 in EU contribution under a Research and Innovation Action (RIA). Licensing terms would need to be negotiated with the coordinator (CERTH, Greece) and relevant consortium partners. As an RIA project, results may have open-access components, but commercial licensing for specific modules is likely required.
Can this work at industrial scale for a large architecture or game studio?+
The project delivered a final system with both frontend applications and backend infrastructure for architecture and video game design platforms. The platform handles web crawling, 3D reconstruction, multilingual text analysis, and linked data for buildings — all pipeline components needed for scale. However, deployment would likely require integration work to match specific studio workflows.
What is the IP situation — can we use this commercially?+
The consortium of 12 partners across 8 countries holds the IP. With 5 SMEs in the consortium, there is clear commercial intent. Specific IP arrangements per module (3D reconstruction, text summarization, crawling tools) would need to be clarified with individual partners who developed each component.
How mature is the 3D reconstruction from video footage?+
The project produced a 'Final demonstrator of 3D reconstruction and model extraction' showing how 3D models are created from video footage, how BIM and GIS data can be extracted, and how this leads to improved 3D representations. This indicates tested technology with working demonstrations on real-world content.
Does it work with our existing design tools (AutoCAD, Revit, Unreal Engine)?+
The project developed BIM and GIS data extraction capabilities and linked data ontologies for building-related resources, which are standard formats used by tools like Revit and AutoCAD. Based on available project data, specific integrations with commercial tools would need verification with the consortium.
What languages does the text analysis support?+
The project delivered a 'Final version of multilingual semantic text analysis' covering multiple languages. The system generates personalized multilingual summaries from reviews, critics, catalogues, and museum guides. Specific supported languages would need to be confirmed with the consortium.
Is there regulatory risk around reusing copyrighted visual content?+
The project specifically addresses content re-purposing from movies, documentaries, paintings, and online art libraries. Based on available project data, the platform works with content providers and crawls online libraries, suggesting licensing and rights management were considered. However, users would need to verify copyright compliance for their specific use cases.
