Fig. 2
From: A richly annotated dataset of co-speech hand gestures across diverse speaker contexts
General data overview. Bar plots show several properties of the dataset including (a). the proportion of different gesture types across the whole dataset, (b). the total number of annotations per speaker, (c). how many unique lexemes each speaker produced, and (d). the percentage of the different gesture types (e.g., beat, iconic) each type of speaker (Lecturer, Politician, Clinician) produced.
