The SHIFT project, launched in October 2022 under the European Commission‘s Horizon Europe program, has released initial versions of its AI-enabled accessibility tools, according to an article on the Europeana cultural site.
The project (in full “MetamorphoSis of cultural Heritage Into augmented hypermedia assets For enhanced accessibiliTy and inclusion”) focuses on the creative and cultural sectors, identified as areas where accessibility and inclusion for people with visual impairments lag behind in the European Union.
Besides accessibility and other technologies, the project combines different disciplines with participants from 13 organizations. These include private companies and accessible technology user associations.
One of these organizations, Romania’s SIMAVI, coordinates the project and is responsible for designing the tools’ architecture, defining functional integrations, and managing end-user pilot validation programs.
Other participating institutions are Germany’s audEERING GmbH (a spin-off of Munich’s Technical University Machine Intelligence and Signal Processing Group), the National Association of Romanian Librarians (ANBPR), the Balkan Museum Network, and the German Federation of the Blind and Partially Sighted (DBSV), among others.
AI Immersive Exhibits for All
Opening new avenues for enhanced digital exhibits — for sighted as well as visually-impaired individuals — SHIFT tools leverage AI, machine learning (ML), haptics, multi-modal data processing, digital content transformation methodologies, semantic representation, linguistic analysis of historical records, and auditory data synthesizers.
The project’s aim is to make historical artifacts and exhibits more accessible, multilingual, and user-friendly. Germany’s audEERING led the development of the SHIFT audio toolkit, which provides audio descriptions and narration for visually impaired exhibit visitors.
The Europeana article lists Albanian, Hungarian, Romanian, Serbian, German, Greek, and English as available audio languages as of May 2025.
According to the article, the SHIFT text-to-speech (TTS) capability in the audio toolkit converts text into high-quality expressive speech using affective speech synthesis, which is capable of conveying emotions like excitement to enhance the narrative. The tool uses 200 “affective” English voices with both native and non-native accents.
The toolkit also features video dubbing and image-to-speech narration. This functionality replaces original video voices with AI-generated speech, including voice cloning, which enables historical figures to “narrate” their own stories.
Besides offering AI voice features, accessible exhibits can now be more immersive by featuring custom background sounds that are period-specific, such as people talking in a group, battlefields, or background music, as shown in a short SHIFT video demo.
These soundscapes are possible through an integration with AudioGen, an AI tool that creates realistic sound environments from text descriptions. The technology also works with augmented reality and virtual reality tours at museums.
Additionally, integrated SHIFT technologies enable the recreation of lost or incomplete historical recordings and the conversion of static images into narrated videos.