Johannes Wagner

Papers

Total Citations

376

H-Index

About

Johannes Wagner is a leading researcher in affective computing and speech processing, whose work has fundamentally advanced the automatic recognition of emotion-related user states from speech. His research focuses on the critical intersection of acoustic and linguistic feature engineering, classification methodologies, and the development of robust systems for human-computer interaction. Wagner’s major contributions include pioneering investigations into which feature types—from low-level descriptors to high-level functionals—are most effective for classifying emotional states, as demonstrated in his highly cited 2007 paper (189 citations) on classifying four emotional user states from children interacting with a pet robot. His 2010 work, “Whodunnit – Searching for the most important feature types signalling emotion-related user states in speech” (146 citations), systematically identified the most salient acoustic and linguistic markers for emotion recognition. Additionally, his 2008 study on patterns and prototypes (41 citations) explored how varying degrees of emotional prototypicality in speech segments affect classification performance. Through these influential studies, Wagner has established foundational knowledge for designing more accurate, feature-efficient emotion recognition systems, making his work essential reading for researchers in speech-based affective computing and intelligent human-machine interfaces.