news
July 23, 2013
updated call for participation
more...
July 23, 2013
updated program
more...
June 24, 2013
updated invited talks
more...
May 27, 2013
acceptance notification date changed (June 17, 2013)
more...

Invited Talks

Artificial Joint Attention: learning, developmental, social, and cultural aspects

Minoru Asada, Ph.D. Professor
Department of Adaptive Machine Systems, Osaka University, Osaka, Japan

Email: asada@ams.eng.osaka-u.ac.jp

Abstract:

Joint attention (hereafter, JA) is one of the important social functions to enable communication between humans. In order to realize this capability in the artificial systems, we have to know how to design it. In this talk, I introduced our studies on JA which have several aspects of joint attention such as learning, developmental, social, and cultural ones. The first one is a bootstrap learning model without any external evaluation. This model is based on the infant inherent capability, that is, visual attention which works with self-evaluation on visual attention. The second one is a developmental learning model with caregiver's evaluation. This model is based on the knowledge that a caregiver can facilitate an infant learning by adjusting the evaluation criterion according to the performance of the infant who has maturational constraint and its advantage. While these two studies have focused on only gaze matching between infant robot and its caregiver, we focus on contingency in interaction between a caregiver and a robot to make the robot acquire various forms of JA behavior. Reproducing social contingency in JA shows open-ended development of JA. The forth one is a study about real infant-caregiver interaction measurement by using kinect sensors. Among various kinds of relationships between them and within one-selves, a developmental change of gaze is observed. Finally, based on these studies, I briefly touch the cultural difference in JA development between east and west.

"Neural representation of visual saliency map in human brain"

Fang Fang, Ph.D. Professor
Department of Psychology, Peking University, Beijing, P.R.China

Email: ffang@pku.edu.cn

Abstract:

The bottom-up contribution to the allocation of exogenous attention is a saliency map, whose neural substrate is hard to identify because of possible contamination by top-down signals. In the first study, we obviated this possibility using stimuli that observers could not perceive, but that nevertheless, through orientation contrast between foreground and background regions, attracted attention to improve a localized visual discrimination. When orientation contrast increased, so did the degree of attraction, and two physiological measures: the amplitude of the earliest (C1) component of the ERP, which is associated with primary visual cortex, and fMRI signals in areas V1-V4 (but not the intraparietal sulcus). Significantly, across observers, the degree of attraction correlated with the C1 amplitude and just the V1 signal. In a second study, we further demonstrated that the response pattern in the first study can generalize to complex natural scenes. These findings strongly support the proposal that a bottom-up saliency map is created in V1, challenging the dominant view that the saliency map is generated in the parietal cortex.