PhD position in Data-Driven Multi-Modal Human-Robot Interaction

One of the most pressing matters that holds back robots from taking on more tasks and reach a widespread deployment in society is their limited ability to understand human communication and take situation-appropriate actions. This PhD position is dedicated to addressing this gap by developing the underlying data-driven models that enable a robot to engage with humans in a socially aware manner.

In order to support long-term interaction, robots need rich models of interaction that include their social context and allow them to act appropriately when faced with uncertainty. Humans use a range of verbal and non-verbal cues to regulate their interactions and are able to adapt how they communicate and cooperate to fit the needs and preferences of their interaction partners. Personalization and adaptation is needed to enable robots to assist people with varying backgrounds and abilities.

The successful applicant will develop machine-learning based and knowledge modelling techniques applied to multi-sensor data (video, audio etc.) of human behaviour that address both the complexity of the multi-modal nature of interactions and uncertainty regarding the sensor data and social situation.The applicant will design and run the experiments to evaluate the created hybrid-AI models through human-robot interaction.

Topics of interest:

1) long-term human-robot interaction

2) affective computing

2) NLP&argument-mining

3) planning & learning for HRI

4) assistive robotics

The Phd candidate will be supervised by Dr. Frank Broz and Prof. Mark Neerincx, focusing on planning and learning for long-term HRI.

To strengthen the social robotics strand of the Interactive Intelligence Group at TU-Delft there are currently two positions available with potential for collaboration.  When you're interested in the other vacancy please have a look at: 

PhD position for Data-Driven Multi-Modal Human-Robot Interaction within DI Lab