This year, Tung Ngo of TU Dublin presented research at two leading international conferences, ACM/IEEE International Conference on Human-Robot Interaction 2026 and CHI Conference on Human Factors in Computing Systems 2026, both widely recognised as premier venues in the fields of human-robot interaction and human-computer interaction.
At HRI 2026, held in March, Tung presented “Vision-Language System using Open-Source LLMs for Consent and Instruction Gestures in Medical Interpreter Robots”. The research introduces a privacy-preserving vision-language framework that enables medical interpreter robots to recognise consent and instruction speech acts and generate appropriate gestures, supporting more effective communication in healthcare settings where language barriers may exist.
At CHI 2026, Tung presented “Qualitative Coding Analysis through Open-Source Large Language Models: A User Study and Design Recommendations”. The paper introduces ChatQDA, an on-device framework powered by open-source large language models designed to support privacy-preserving qualitative data analysis. Findings from a mixed-methods user study showed that participants viewed the system as highly usable and efficient, while expressing “conditional trust” in its outputs, particularly regarding the interpretation and consistency of qualitative coding.
Together, the two papers highlight the potential of open-source AI technologies to address privacy, accessibility and trust challenges in both healthcare and research contexts.
