In order to learn the mappings from words to referents, children must integrate co-occurrence information across individually ambiguous pairs of scenes and utterances, a challenge known as cross-situational word learning. In machine learning, recent multimodal neural networks have been shown to learn meaningful visual-linguistic mappings from…

Assessed word-learning skills of 19 deaf/hard-of-hearing preschoolers either with novel mapping strategy to learn new words, or after minimal exposure when reference was explicitly established. Found that 11 children were able to learn words in both contexts, 5 only in the second, and 2 in neither. The latter 7 children eventually were able to…