Audiovisual Learning Why realistic lip movements might not make social robots better for language learning Child Speech Recognition Automatic Speech Recognition finally works well for young children. Enjoyment Prediction Large Language Models can predict how much a user will enjoy a human-robot dialog before seeing the user's response. Expressive Furhat LLMs can generate real-time custom facial expressions Generative AI for Social Robots in Education Generative AI models will finally enable autonomous and adaptive social robots in education Human Alignment To LLMs Humans change their social behaviour to LLMs. Miscommunication Detection Why robots are bad at detecting their mistakes Multimodal Social Conversations Using Vision-Language Models for Human-Robot Social Dialogue PhD Thesis Visually Grounded Adaptive Social Dialogue for Robots L2 Speaking Proficiency Assessment Automatically detecting speaking proficiency in a second language. Visual Conversation Starters Generating visually-grounded conversation-starting questions for Human-Robot Interaction