Catch me if you can

Abstract

Embodied conversational agents become more and more realistic concerning their conversational and their nonverbal behaviors. But if the information conveyed nonverbally exhibits clues that are not consistent with the verbal part of an agent's action, how will the user react to such a discrepancy? Masking ones real emotions with a smile is a naturally occuring example of such a discrepancy. But such masks are often deficient and thus subtle clues of lying and deceiving manifest themselves in facial expressions. The questions is how users will react to these clues if they are conveyed by an agent. Will they render an application unattractive or on the contrary more human-like? In this paper, we examine such facial clues to deception and present the results of two empirical studies: i.) lies in monologues by a talking head presenting movies, ii.) lies in an interactive game of dice.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 80 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.