NET1: How to Research

On the first day of the workshop the attendees were greeted by the coordinator Stefan Wermter with an introduction to the TRAIL project. The introduction was followed by the presentations of the doctoral candidates (DCs) based on their research proposals drafted during the first learning circle activity (LCA).

• Sergio Lanza (UHAM), "Explaining Robot Behavior through Concept Bottleneck Models"
• Julia Gachot (UHAM), "Explainable Semantic and Pragmatic Representations for Speech Recognition"
• Tien Pham (UMAN), "Interactive Language Explanations for Trust in Human-Robot Communication"
• Theodor Wulff (UMAN), "Transparent and Grounded Communication with Abstract Language"
• Pradip Pramanick (UNINA), "Proactive and Multimodal Explanatory Robot Behavior from Contextual Cues"
• Tamlin Love (CSIC), "Fostering Human-Robot Mutual Understanding by Explaining the Internal Beliefs"
• Ferran Gebellí (PAL), "Design and Implementation of an Ethical Black Box for Service Robots"

External speakers were Michael Fisher with a talk on "Sustainability in AI", and Gavin Brown with a talk on "How to do a PhD". The evening activities were a Managment and Supervisory Board Meeting in parallel to a Team Building Session for the DCs.

The second day started with a talk by Igor Farkaš on "Efficient Dissemination and Scientific Publishing", followed by a talk by TRAIL's ethical advisor Alan Winfield on "Ethics in AI".  Angelo Cangelosi and Silvia Rossi held presentations on "Research Methods" and "Design of HRI Experiments" respectively.

Alessandro Sperduti introduced the second LCA and allocated the DCs into their new learning circles.

In the evening the DCs participated in two labs introducing them to software tools in robotics (ROS, Aldebaran middleware) and neural networks (PyTorch, Hugging Face, Nvidia Nemo).

See you soon in Barcelona!