Effects of forensically-realistic facial concealment on auditory-visual consonant recognition in quiet and noise conditions

The study presented in this paper investigates auditory-only and auditory-visual (AV) consonant recognition where the talker’s face is obscured by various types of face-concealing garments and headgear. Observers’ consonant identification performance across the various ‘facewear’ conditions was tested both in quiet listening conditions (Experiment 1), and when the speech stimuli were embedded in 8-talker babble noise (Experiment 2). Statistical analysis of the responses collected from 82 phonetically-untrained subjects (N = 43, quiet; N = 39, noise) revealed a significant AV effect in both experiments. However, the strength of the effect varied considerably as a function of facewear type. The findings are discussed in the context of previous research on AV speech perception, which aims to identify the facial regions that are particularly important for the extraction of visual speech cues.