A constructive model for the development of joint attention

Abstract

This paper presents a constructive model by which a robot acquires the ability of joint attention with a human caregiver based on its embedded mechanisms of visual attention and learning with self-evaluation. The former is to look at a salient object in the robot's view, and the latter is to learn sensorimotor co-ordination when visual attention has succeeded. Since the success of visual attention does not always correspond to the success of joint attention, the robot has incorrect learning data for joint attention as well as correct data. However, the robot is expected statistically to lose incorrect data as outliers since such data do not have any correlation in the sensorimotor co-ordination while correct data have a correlation. The robot consequently acquires the ability of joint attention by finding the correlation in the sensorimotor co-ordination even if multiple objects are placed at random positions in an environment and a human caregiver does not provide any task evaluation to the robot. The experimental results show that the proposed model makes the robot reproduce the developmental process of infants' joint attention. Therefore, the proposed model could be one of the models to explain how infants develop the ability of joint attention.

Autonomous Development of Social Referencing Skills

“…Here, we don't solve the question of joint attention which is an social referencing skill. Joint attention may also be reached using a learning protocol similar to Nagai [18] (developmental model for the joint attention). We think this approach can provide new interesting insights about how humans can develop social referencing capabilities from sensorimotors dynamics.…”

Learning to Interpret Pointing Gestures: Experiments with Four-Legged Autonomous Robots

“…It has been tested in the case of gaze following in the context of a human-robot interaction by Nagai et al [8] and found to be feasible. This hypothesis relies on the assumption that the correlation is sufficiently strong to be discovered in practice.…”

“…To the best of our knowledge, this is the first experiment on pointing and pointing gesture recognition using two robots. Earlier robotic experiments focused on gaze and pointing gesture recognition between a human and a robot [8,9]. A computational model of reward-based emergence of gaze following was presented in Carlson and Triesch [10].…”

Learning of Joint Attention from Detecting Causality Based on Transfer Entropy

“…A number of robotics researchers have pointed out that it may play an important role in enabling smooth communication between humans and robots [2], [3]. Synthetic studies have been also conducted to understand the underlying mechanisms of the development of joint attention [4], [5]. We can expect such studies to enable realizing adaptive robot to changes in an environment including humans as well as to help understanding of human development [6], [7].…”