Interactive Language Learning by Robots: The Transition from Babbling to Word Forms

Abstract

The advent of humanoid robots has enabled a new approach to investigating the acquisition of language, and we report on the development of robots able to acquire rudimentary linguistic skills. Our work focuses on early stages analogous to some characteristics of a human child of about 6 to 14 months, the transition from babbling to first word forms. We investigate one mechanism among many that may contribute to this process, a key factor being the sensitivity of learners to the statistical distribution of linguistic elements. As well as being necessary for learning word meanings, the acquisition of anchor word forms facilitates the segmentation of an acoustic stream through other mechanisms. In our experiments some salient one-syllable word forms are learnt by a humanoid robot in real-time interactions with naive participants. Words emerge from random syllabic babble through a learning process based on a dialogue between the robot and the human participant, whose speech is perceived by the robot as a stream of phonemes. Numerous ways of representing the speech as syllabic segments are possible. Furthermore, the pronunciation of many words in spontaneous speech is variable. However, in line with research elsewhere, we observe that salient content words are more likely than function words to have consistent canonical representations; thus their relative frequency increases, as does their influence on the learner. Variable pronunciation may contribute to early word form acquisition. The importance of contingent interaction in real-time between teacher and learner is reflected by a reinforcement process, with variable success. The examination of individual cases may be more informative than group results. Nevertheless, word forms are usually produced by the robot after a few minutes of dialogue, employing a simple, real-time, frequency dependent mechanism. This work shows the potential of human-robot interaction systems in studies of the dynamics of early language acquisition.

Employing Humanoid Robots for Teaching English Language in Iranian Junior High-Schools

“…This¯nding is consistent with those of prior studies using RALL on general English pro¯ciency or similar studies. 3,13,14,[28][29][30][31][32][33] However, this study is more in line with Hyun et al 29 who worked with preschoolers' vocabulary using RALL and found their improvement was signi¯cantly better than the CALL group.…”

See You See Me

“…First, the task was inspired from similar developmental psychology experiments showing how young children learn to associate visual objects with novel words through child-parent social interactions [Baldwin 1993; Estes, Evans, Alibali and Saffran 2007; Yu and Smith 2012]. Second, there is a trend in developmental robotics that users teach robots human languages through human-robot interactions [Lyon, Nehaniv and Saunders 2012; Marocco, Cangelosi, Fischer and Belpaeme 2010; Tanaka and Matsuzoe 2012; Xu, et al 2013; Yu, et al 2010]. Moreover, this joint task naturally engaged participants to interact with the robot without any constraint on what they had to do or what they had to say.…”

Beyond Vision: Extending the Scope of a Sensorimotor Account of Perception

“…In this case the hearer can often have a passive as well as an active role. Active listening occurs when a subject directs their attention to speech or other sounds, so humanoid robots acquiring linguistic skills through interaction with a human will typically be listening in an active mode [17]. On the other hand, auditory perception can be passive: a sleeping person can be woken by sounds, and some loud or incessant noise can be an unpleasant experience that the hearer cannot escape.…”

“…This highlights the point that orthographic transcripts of spontaneous speech may not be a close match to perceived sounds -a point that needs to be taken into account when research into, for instance, child language acquisition uses corpora of written transcripts of child directed speech. The observed fact that content words are more likely to have consistent phonemic structure is exploited in work simulating the learning of word forms by a humanoid robot which interacts with humans [17].…”