Machine Learning & Data Analytics

Our results indicate that the considered acoustic features are related to engagement levels for both the child-child and child-robot interaction.

The two case studies presented here might, at first glance, seem to represent very different points in that design space, but they are highly related with respect to the turn-taking problems and challenges they expose.

In this paper, using different formant-related measurements as exemplary analysis features generated within articulatory-phonetic guidelines, we demonstrate the nonlinear relationships of children’s physical parameters to their voice.

This paper highlights the issues with keyword spotting using a simple two-word game played by children of different age groups and gives quantitative performance assessments using a novel keyword spotting technique that is especially suited to such scenarios.

We describe phone recognition experiments on hand labelled data for children aged between 5 and 9.

We report results for an online multi-keyword spotter in a game that contains overlapping speech, off-task side talk, and keyword forms that vary in completeness and duration.

We propose a neural model to learn argument embeddings from the context by explicitly incorporating dependency relations as multiplicative factors, which bias argument embeddings according to their dependency roles.

We explore what happens in the increasingly likely situation that a robot has sensed information about a child of which the child is unaware, then discloses that information in conversation in an effort to personalize the child’s experience.

In this paper, we first rigorously compare the two algorithms and in the process develop several extensions, including a version of EBP for continuous regression problems and a PBP variant for binary classification.

We showcase the efficacy of our approach in a user study, where we demonstrate orders-of-magnitude improvements in search quality compared to baseline systems.