Funded by the Fonds National de la Recherche Luxembourg in the framework of their partnership block grant programme IPBG, the University of Luxembourg has created the COLLABORATION 21 (C21) programme. It teams up researchers from the University with Cisco – the world leader in collaboration technologies and SCRIPT – Luxembourg’s national agency for innovation and digitalisation in schools. Together, they propose a highly interdisciplinary programme encompassing seven research challenges at the intersection of collaborative skills and technologies to support education and the workforce.

Machine Learning & Arts: The Smart Photo Booth

The goal of this project is to develop a playful and interactive intelligent machine – the Smart Photo Booth – where the users can experiment with AI, and learn about the process of how intelligent machines are trained.


Artificial Intelligence (AI) is more and more present on our daily lives (often without people noticing it). There are many misconceptions about AI, and as AI researchers, it is our responsibility to initiate a dialogue with the public to discuss what AI is, where the research is going and the ethical issues it raises.

Within this project, we developed a playful and interactive Smart Photo Booth, the robot CUTIE and combined AI and Art to engage different audiences and entice them to explore computational methods.

The idea behind the Smart Photo Both is to apply AI to manipulate images to resemble artistic styles for two reasons: i) humans are very visual and drawn to images, which are a powerful media to provoke interaction and enable easy communication; ii) image manipulation and deep fakes are a good entry point to discuss the potential of AI.

Thus, the Smart Photo Booth enabled us to communicate the possibilities brought by AI and machine-learning to a wider audience, in particular teenagers, including those who are not technology drawn.

The Smart Photo Booth, CUTIE, is a QT Robot ( that was tailored to take pictures and use them as input to create new images as if they were painted by a renowned artist. The software makes a “style transfer”, using Generative Adversarial Networks (GAN), a design of neural networks capable of generating images and sound. The development of CUTIE was done in partnership with LuxAI, a spin-off of the University of Luxembourg. To accompany CUTIE we developed a video, in collaboration with an Art History expert, to convey to the audience a brief history of portraits and selfies and explain the machine learning and AI concepts behind CUTIE.

We worked together with the Scienteens Lab of the University of Luxembourg ( to deliver workshops in machine learning & art for high school students. During these workshops, the students became acquainted with the main concepts of artificial intelligence, machine learning, and the style transfer algorithm. Participants were provided code, which they could experiment with. Furthermore, the technologies were placed in a larger framework and multiple examples were discussed on how it affects our daily lives. The upside of these technologies using deep neural networks was debated as well as the consequences and dangers of the possibility to generate high-quality images (e.g. deep fakes). In addition, this project was a pilot for a joint PSP-Flagship Proposal – BeCoS: Become a Computer Scientist, led by the Scienteens Lab. BeCoS aims to promote computer literacy to encourage computer science as a career path for teenagers and especially for young women. It is now part of the Computer Science program of the Scienteens Lab (

We also collaborated with the Luxembourg Science Center (LSC). CUTIE went to the LSC and interacted with around 100 visitors over three Saturdays. Researchers from the University of Luxembourg accompanied CUTIE and facilitated the interaction with the public. The participants really enjoyed the interaction with the robot and the opportunity to choose different art styles.

In 2022, the Smart Photo Booth was installed in the Computational Creativity Hub at the University of Luxembourg and integrated with the AI&Art Project of the at the Esch2022.


This project is a collaboration with the:

Supported by the Luxembourg National Research Fund (FNR) PSP-Classic Project 15417971

Within the context of the Esch2022 European capital of culture, our project proposal for an AI & Art pavilion has been accepted.

The project is cofounded by Esch22 and the university of Luxembourg. PI: Prof. Leon van der Torre. Therefore, the AI Robolab will be responsible for the organization of an AI & Art pavilion that will be included in the activates of Esch 2022 European Capital of Culture.

The pavilion involves several corners and projects developed by artists and researchers interested in creative technologies as well as staff from the AI Robolab. The kickoff meeting of this project was organized by the AI Robolab and held on 25th of September 2020.

Project EXPECTATION (2021-2024) Accepted.

EXPECTATION is s CHIST-ERA (ERA-NET and FET supported project) on eXplainable AI (XAI) entitled: Personalized Explainable Artificial Intelligence for decentralized agents with heterogeneous knowledge.


The project involves 4 partners:

  • University of Luxembourg (PI: Prof. Leon van der Torre),
  • HES-SO, University of Applied Sciences and Arts Western Switzerland,
  • Alma Mater Studiorum Università di Bologna, Italy and
  • Özyeğin University, Turkey.

Project description

Explainable AI (XAI) has recently emerged proposing a set of techniques attempting to explain machine learning (ML) models. The recipients (explainee) are intended to be humans or other intelligent virtual entities. Transparency, trust, and debuging are the underlying features calling for XAI. However, in real-world settings, systems are distributed, data are heterogeneous, the “system” knowledge is bounded, and privacy concerns are subject to variable constraints. Current XAI approaches cannot cope with such requirements. Therefore, there is a need for personalized explainable artificial intelligence. We plan to develop models and mechanisms to reconcile sub-symbolic, symbolic, and semantic representations leveraging on the agent-based paradigm. In particular, the proposed approach combines inter-agent, intra-agent, and human-agent interactions to benefit from both the specialization of ML agents and the establishment of agent collaboration mechanisms, which will integrate heterogeneous knowledge/explanations extracted from efficient black-box AI agents. The project includes the validation of the personalization and heterogeneous knowledge integration approach through a prototype application in the domain of food and nutrition monitoring and recommendation, including the evaluation of agent-human explainability, and the performance of the employed techniques in a collaborative AI environment.