Multimodal Person Evaluation: First Impressions from Faces, Voices, and Names

Mila Mileva

This interactive plot shows how different first impression traits based on face, voice, and name cues are related to one another in meaning based on behavioural data. Increasing the number on the slider below shows words that a more closely related to one another based on the data. Clustering with word similarity of between 0.4 - 0.6 produces the most interpretable pattern of results. Hovering on each point will reveal the word that it represents. Within each cluster, points that are located closer to one another are more closely related.

Trait Clustering - Face+Voice+Name