Decades of research has studied how language learning infants learn to discriminate speech sounds, segment words, and associate words with their meanings. While gradual development of such capabilities is unquestionable, the exact nature of these skills and the underlying mental representations yet remains unclear. In parallel, computational studies have shown that basic comprehension of speech can be achieved by statistical learning between speech and concurrent referentially ambiguous visual input. These models can operate without prior linguistic knowledge such as representations of linguistic units, and without learning mechanisms specifically targeted at such units. This has raised the question whether knowledge of linguistic units, such as phone(me)s, syllables, and words, could actually emerge as latent representations supporting the translation between speech and representations in other modalities, and instead of the units ever being proximal learning goals for the learner. In this study, formulate this idea as the so-called latent language hypothesis (LLH), connecting linguistic representation learning to general predictive processing within and across sensory modalities. We review the extent that the audiovisual aspect of LLH is supported by the existing computational studies. We then explore LLH further in extensive learning simulations with different neural network models for audiovisual cross-situational learning, and comparing learning from both synthetic and real speech data. We investigate whether the latent representations learned by the networks reflect phonetic, syllabic, or lexical structure of input speech by utilizing an array of complementary evaluation metrics related to linguistic selectivity and temporal characteristics of the representations. As a result, we find that representations associated with phonetic, syllabic, and lexical units of speech indeed emerge from the audiovisual learning process. The finding is also robust against variations in model architecture or characteristics of model training and testing data. The results suggest that cross-modal and cross-situational learning may, in principle, assist in early language development much beyond just enabling association of acoustic word forms to their referential meanings.

Earlier research has suggested that human infants might use statistical dependencies between speech and non-linguistic multimodal input to bootstrap their language learning before they know how to segment words from running speech. However, feasibility of this hypothesis in terms of real-world infant experiences has remained unclear. This paper presents a step towards a more realistic test of the multimodal bootstrapping hypothesis by describing a neural network model that can learn word segments and their meanings from referentially ambiguous acoustic input. The model is tested on recordings of real infant-caregiver interactions using utterancelevel labels for concrete visual objects that were attended by the infant when caregiver spoke an utterance containing the name of the object, and using random visual labels for utterances during absence of attention. The results show that beginnings of lexical knowledge may indeed emerge from individually ambiguous learning scenarios. In addition, the hidden layers of the network show gradually increasing selectivity to phonetic categories as a function of layer depth, resembling models trained for phone recognition in a supervised manner.

No abstract

No abstract

Previous computational models of early language acquisition have shown how linguistic structure of speech can be acquired using auditory or audiovisual learning mechanisms. However, real infants have sustained access to both uni- and multimodal sensory experiences. Therefore, it is of interest how the uni- and multimodal learning mechanisms could operate in concert, and how their interplay might affect the acquisition dynamics of different linguistic representations. This paper explores these questions with a computational model capable of simultaneous auditory and audiovisual learning from speech and images. We study how the model’s latent representations reflect phonemic, lexical, and semantic knowledge as a function of language experience. We also test how the findings vary with differential emphasis on the two learning mechanisms. As a result, we find phonemic learning always starting to emerge before lexical learning, followed by semantics. However, there is also notable overlap in their development. The same pattern emerges irrespectively of the emphasis on auditory or audiovisual learning. The result illustrates how the acquisition dynamics of linguistic representations are decoupled from the primary learning objectives (mechanisms) of the learner, and how the emergence of phonemes and words can be facilitated by both auditory and audiovisual learning in a synergetic manner.