We address prediction of turn-taking considering related behaviors such as backchannels and fillers. Backchannels are used by the listeners to acknowledge that the current speaker can hold the turn. On the other hand, fillers are used by the prospective speakers to indicate a will to take a turn. We propose a turntaking model based on multitask learning in conjunction with prediction of backchannels and fillers. The multitask learning of LSTM neural networks shared by these tasks allows for efficient and generalized learning, and thus improves prediction accuracy. Evaluations with two kinds of dialogue corpora of human-robot interaction demonstrate that the proposed multitask learning scheme outperforms the conventional single-task learning.

In human-human dialogue, especially in attentive listening such as counseling, backchannels are important not only for smooth communication but also for establishing rapport. Despite several studies on when to backchannel, most of the current spoken dialogue systems generate the same pattern of backchannels, giving monotonous impressions to users. In this work, we investigate generation of a variety of backchannel forms according to the dialogue context. We first show the feasibility of choosing appropriate backchannel forms based on machine learning, and the synergy of using linguistic and prosodic features. For generation of backchannels, a framework based on a set of binary classifiers is adopted to effectively make a "not-to-generate" decision. The proposed model achieved better prediction accuracy than a baseline which always outputs the same backchannel form and another baseline which randomly generates backchannels. Finally, evaluations by human subjects demonstrate that the proposed method generates backchannels as naturally as human choices, giving impressions of understanding and empathy.

Attentive listening systems are designed to let people, especially senior people, keep talking to maintain communication ability and mental health. This paper addresses key components of an attentive listening system which encourages users to talk smoothly. First, we introduce continuous prediction of end-of-utterances and generation of backchannels, rather than generating backchannels after end-point detection of utterances. This improves subjective evaluations of backchannels. Second, we propose an effective statement response mechanism which detects focus words and responds in the form of a question or partial repeat. This can be applied to any statement. Moreover, a flexible turn-taking mechanism is designed which uses backchannels or fillers when the turnswitch is ambiguous. These techniques are integrated into a humanoid robot to conduct attentive listening. We test the feasibility of the system in a pilot experiment and show that it can produce coherent dialogues during conversation.

No abstract