Séminaire : Which visual and auditory cues are used in time-to-collision judgments?

Daniel OBERFELD (PhD TU Berlin, HDR U Mainz) fera une présentation de ses travaux de recherche le vendredi 30 Octobre 2015 à 10h30 dans la salle des séminaires, 4 rue Boussingault, au 1er étage. Le séminaire s'adresse à un public large et sera suivi d'une discussion autour des modèles (physiques et autres) qui visent à estimer les temps de collision. Les doctorants pourront faire valider une feuille de présence.

Titre : Which visual and auditory cues are used in time-to-collision judgments?

Équipe : MMB

Résumé : Judging when a moving object will arrive at a designated position in space is crucial in traffic and in sports. Previous studies of time-to-collision (TTC) estimates focused on the effectiveness of visual TTC information (e.g., visual "tau" provided by the optical expansion pattern, disparity). Fewer studies measured the integration of auditory and visual TTC information. The latter studies did not systematically investigate the use of "heuristic" cues such as final optical size, and in general did not provide compelling information about the relative weighting of different auditory and visual cues. I will present a study in which participants judged the TTC of an approaching object presented in the visual or auditory modality, or both concurrently. A psychophysical reverse correlation approach was used to estimate the influence of auditory and visual cues on TTC estimates. Experiment 1 presented simple lab-type stimuli. Multiple regression analyses showed that TTC estimates were based on both auditory and visual information, and that visual information was weighted more heavily. In the auditory modality, the TTC estimates were more strongly influenced by the final sound pressure level than by the actual auditory TTC. Thus, we found an analogy of the well-documented effect of final optical size on visual TTC estimates. In experiment 2, we presented a higher fidelity visual simulation showing an urban street scenario, combined with a comparably simple auditory simulation of an approaching vehicle. The data confirmed the use of heuristic cues in both the auditory and the visual domain. In the audio-visual condition, the estimates were dominated by visual cues, most likely due to the higher fidelity of the visual compared to the auditory simulation. I will discuss research perspectives and potential implications for traffic safety.

Bio : Daniel OBERFELD is Associate Professor of Psychology in the Department of Experimental Psychology at Johannes Gutenberg-Universität in Mainz, Germany. He completed his Ph.D. in Psychology at the Technische Universität Berlin in 2005, and his Habilitation in Psychology at the Johannes Gutenberg-Universität Mainz in 2011. His research topics comprise auditory, visual, and multimodal perception, emotion, and psychophysical methods and models