![HER[AI]TAGE](https://her-ai-tage.pou-cakovec.hr/wp-content/uploads/2025/03/logo-1.png){kind=logo}
Your journey through the course: A recap
Congratulations on completing this conceptual journey through the world of AI in cultural heritage preservation. You have explored a comprehensive framework, starting with the foundational principles of a heritage project as modelled by the HER[AI]TAGE project.
You followed a complete project lifecycle: from the crucial first steps of gathering stories and digitising artefacts using best practices, to navigating the complex ethical landscape and securing informed consent under GDPR and UNESCO principles. You saw how AI can act as a powerful assistant in processing transcribed texts – summarising, translating, and extracting keywords – and in drafting rich descriptive texts and narrative scripts. You then learned how to assemble these assets into a coherent and accessible digital exhibit and the vital importance of engaging with the community for feedback in a collaborative loop.
Building on that foundation, the subsequent modules explored more advanced creative applications. You discovered how AI can help reimagine the past by visually reconstructing damaged historical images, how it can give new voice to old words by transforming traditional poems into contemporary songs, and how it can bring stories to life by generating evocative video clips from oral histories. This course has equipped you with a conceptual toolkit to not only preserve heritage but to creatively and ethically engage with it.
The future is now: Advanced AI applications in cultural heritage
The applications we have explored are just the beginning. The field of AI is evolving at a breathtaking pace, opening new frontiers for preserving, understanding, and interacting with our shared heritage.
The following table summarises some of the advanced ways AI is already being used in the sector.
| Application area | Core AI technology | Real-world example / case study |
| 3D reconstruction of artefacts | Photogrammetry, Neural Radiance Fields (NeRFs), 3D Gaussian Splatting (3DGS) – Techniques that allow AI to build detailed 3D models from 2D images | Reconstructing the detailed reliefs of Indonesia’s Borobudur Temple from single, flat historical photographs with 95% accuracy. |
| Immersive AR/VR experiences | Augmented Reality (AR) & Virtual Reality (VR) engines, AI-driven content generation | The SHELeadersVR project uses VR in museums and AR at archaeological sites to tell the stories of female rulers in the Balkans. |
| Custom LLMs for archival research | Fine-tuning of open-source LLMs (e.g., Llama, Gemma) on specific project data, creating an expert search tool for that collection | Building a smart search system for a digital archive that allows users to ask natural language questions and receive nuanced, context-aware answers drawn directly from the source material. |
| Predictive conservation | Predictive analytics, machine learning on sensor data | The AI-powered Heritage Building Information Modelling (HBIM) system used in Venice analyses data to predict salt damage and structural fatigue in historic buildings, enabling proactive maintenance. |
| AI-assisted provenance research | Pattern recognition, data mining on historical documents | Analysing vast datasets of auction records, letters, and shipping manifests to trace the ownership history (provenance) of cultural artefacts, helping to authenticate items and resolve ownership disputes. |
A final word of encouragement
The most important step is the next one you take. Use the knowledge from this course to start a small project, even if it is just recording one story from a family member. Use AI as your assistant to help you plan, process, and present it. Learn by doing. Iterate, refine, and grow. The work of preserving heritage is a marathon, not a sprint, and every story you save adds a priceless thread to the rich tapestry of human experience.