If traditional services (loans, consultation, etc.) remain at the heart of the activities of the Cantonal University Library of the Canton of Fribourg (BCU), enhancing and sharing the digital heritage presents new challenges for libraries. In particular, how to rethink the representation of heritage in the digital era? How to create a relationship between physical documents and online access, and which role a digital collection plays on social dynamics and collective identity? While considerable efforts have been made by many institutions (libraries, museums, foundations, etc.) to digitize their collections, little work has been done to explore the potential for the valorization of these digitized heritages, apart from remote access and research tools.
Taking advantage of the creation of a new building for the library of Fribourg, the EPFL+ECAL Lab and the BCU established a design research partnership to address the above-mentioned research questions.
To increase the attractiveness of the collection, previous research has shown the need to create identification with the content. Through qualitative user testing, we were able to develop a dedicated strategy for BCU content: to create identification and appeal, content displaying and/or describing scenes from everyday life should be used in combination with an interaction system underlying a specific part of the content, i.e., an object, face, sign, etc.
The challenge and the main topic of the proposed project here are to explore algorithms capable of captioning some parts of the visual content and creating semantic links between similar content. The ultimate goal of the overall project is to create an interactive installation.
The involved student will be part of the EPFL+ECAL Lab team conducting the design research project composed of engineers, designers, and psychologists. The final overall project will be displayed in the new building of the BCU in Fribourg.
- Machine learning
- Image captioning
- Some knowledge of machine learning, deep-learning.
- Basics of computer vision.
- Proficiency in Python.