Norwegian engineers have recorded several different voices of adults who uttered more than 1000 phrases, on which was built dictionary of sounds and words. Sound tracks were mixed and combined the synthetic voice. In order to make it sound like a child, the experts carried out a comparative computer analysis of this mixed and children’s voices. Catching the differences, the engineers tried to improve the resulting synthetic child’s voice and make it more believable.
However, it sounds a bit strange. But the developers are confident that mutual understanding between adults and children is worth in order to teach dumb kids use the new device.