‘EuropeanaTech x AI’ is a webinar series coming on 23 April that will explore both the benefits and challenges of Artificial Intelligence and Machine Learning applications on cultural heritage. Each session will present a topic or project from the sector.
Cultural Heritage is at the forefront of emergent innovations and applications of AI technology. While still not widely used several projects have adopted machine learning, deep learning and explainable AI to foster research, presentation and access to heritage practices.
As our understanding of AI grows, so does the framework and uses, with the ethics and benefits debated in academia. AI can shift through and generate tons of data pertinent to heritage collections. Moreover, AI can identify and conceptualise data with uses as:
- Automatic identification
- Decipher ancient languages
- Restore ancient texts
- Decode epigraphic marks
- Detect undiscovered archaeological sites
- Detect threats and impact of climate change
- Automatic digitisation and 3D printing
Explainable AI (XAI) applications make art and cultural heritage accessible and inclusive to wider audiences, as its targets are non-tech savvy. XAI can present information in multiple formats and languages, thus offering services designed to address specific disabilities or minorities. Additionally, the presentation method can accommodate an interactive narrative experience to the user through storytelling or VR. Image-captioning is key to enhancing the visitor experience and learning as XAI can identify points between an image and a caption.
There already are in place tailored visitor guides such as for people with mobility issues. Let us briefly introduce you to some notable cultural projects until we delve deeper in our coming posts:
MonuMAI is a mobile app that learns and recognises architectural styles from photographs. Developed in Grenada, citizens and researchers use Deep Learning to enhance their heritage.
INSIGHT (Intelligent Neural Systems as InteGrated Heritage Tools) a research project aiming to enrich collections with metadata with a focus on language. Their efforts will make available multilingual and cohesive descriptions and information across cultural institutions.
Time Machine is a European project that maps the social, cultural and geographical evolution over time with a digital information system. Moreover, beyond data gathering, they plan to reconstruct 4D simulations with Augmented/Virtual Reality (AR/VR) applications.
VizWiz addresses artificial intelligence challenges and built datasets from blind users. Their efforts hope to educate and encourage professionals and researchers to develop assistive technologies. They collect data by blind users of a mobile app by taking a picture and verbalising a question about it. So far, their work tackles Image Captioning, Image Quality, Visual Q&A and more.
Lazarillo Project shares a similar vision through “ImageNet” for the blind, a dataset of images gathered from phones and wearable cameras of visually impaired users using an app. Their work will improve Image Captioning and Q&A and facilitate their daily lives.
Copernicus for Cultural Heritage is a European Earth observation programme, using Earth Observation (EO) data to preserve and manage cultural heritage sites. Daily heritage management activities better through the integration of EO Data, such as:
- Land-use change maps
- Natural subsidence, ground motion detection
- Risk assessment maps
- Archaeological sites monitoring and identification (e.g. buried sites)
- Monitoring of the destruction or looting of sites
- Urban sprawl monitoring
- Climate Change indicators
- Air pollution monitoring
- Coastline monitoring (erosion)
Microsoft’s AI for Cultural Heritage works with non-for-profits and cultural institutions to preserve cultural heritage through AI use, with several projects already into fruition.
IVOW uses conversational AI and storytelling to help clients create compelling narratives. A beautiful example of cultural engagement is the Indigenous Knowledge Graph (IKG), forming a structure around stories and traditions or “an AI agent who offers you recipes and insights about under-represented people, cultures, and perspectives, including ancient traditions, and vibrant food cultures from around the world.”
AI technology is at its frontier with endless capabilities and applications. In culture, AI already helps gather, process and interpret massive amounts of data, thus enhancing the quality of research and accuracy of information. Working with AI in art and heritage, we have the potential for greater understanding, education, tailored and specialised access, and even develop services we have not yet imagined.
I want to learn more:
Accessible Cultural Heritage through Explainable Artificial Intelligence by Natalia Díaz-Rodríguez and Galena Pisoni.
PS:”just shower thoughts” imagine Skynet arriving and instead of the events of Terminator they start planning and managing sites and museums.