Infants have been shown to rely both on auditory and visual cues when processing speech. We investigated the impact of COVID-related changes, in particular of face masks, in early word segmentation abilities. Following up on our previous study demonstrating that, by 4 months, infants already segmented targets presented auditorily at utterance-edge position, and, using the same visual familiarization paradigm, 7–9-month-old infants performed an auditory and an audiovisual word segmentation experiment in two conditions: without and with an (...) FFP2 face mask. Analysis of acoustic and visual cues showed changes in face-masked speech affecting the amount, weight, and location of cues. Utterance-edge position displayed more salient cues than utterance-medial position, but the cues were attenuated in face-masked speech. Results revealed no evidence for segmentation, not even at edge position, regardless of mask condition and auditory or visual speech presentation. However, in the audiovisual experiment, infants attended more to the screen during the test trials when familiarized with without mask speech. Also, the infants attended more to the mouth and less to the eyes in without mask than with mask. In addition, evidence for an advantage of the utterance-edge position in emerging segmentation abilities was found. Thus, audiovisual information provided some support to developing word segmentation. We compared 7–9-monthers segmentation ability observed in the Butler and Frota pre-COVID study with the current auditory without mask data. Mean looking time for edge was significantly higher than unfamiliar in the pre-COVID study only. Measures of cognitive and language development obtained with the CSBS scales showed that the infants of the current study scored significantly lower than the same-age infants from the CSBS normative data. Our results suggest an overall effect of the pandemic on early segmentation abilities and language development, calling for longitudinal studies to determine how development proceeds. (shrink)

Statistical learning is often taken to lie at the heart of many cognitive tasks, including the acquisition of language. One particular task in which probabilistic models have achieved considerable success is the segmentation of speech into words. However, these models have mostly been tested against English data, and as a result little is known about how a statistical learning mechanism copes with input regularities that arise from the structural properties of different languages. This study focuses on statistical word (...) segmentation in Arabic, a Semitic language in which words are built around consonantal roots. We hypothesize that segmentation in such languages is facilitated by tracking consonant distributions independently from intervening vowels. Previous studies have shown that human learners can track consonant probabilities across intervening vowels in artificial languages, but it is unknown to what extent this ability would be beneficial in the segmentation of natural language. We assessed the performance of a Bayesian segmentation model on English and Arabic, comparing consonant-only representations with full representations. In addition, we examined to what extent structurally different proto-lexicons reflect adult language. The results suggest that for a child learning a Semitic language, separating consonants from vowels is beneficial for segmentation. These findings indicate that probabilistic models require appropriate linguistic representations in order to effectively meet the challenges of language acquisition. (shrink)

We present a Chinese word segmentation system submitted to the closed track of Sighan bakeoff 2005. Our segmenter was built using a conditional random field sequence model that provides a framework to use a large number of linguistic features such as character identity, morphological and character reduplication features. Because our morphological features were extracted from the training corpora automatically, our system was not biased toward any particular variety of Mandarin. Thus, our system does not overfit the variety of (...) Mandarin most familiar to the system's designers. Our final system achieved a F-score of.. (shrink)

The informativity of a computational model of language acquisition is directly related to how closely it approximates the actual acquisition task, sometimes referred to as the model's cognitive plausibility. We suggest that though every computational model necessarily idealizes the modeled task, an informative language acquisition model can aim to be cognitively plausible in multiple ways. We discuss these cognitive plausibility checkpoints generally and then apply them to a case study in word segmentation, investigating a promising Bayesian segmentation (...) strategy. We incorporate cognitive plausibility by using an age-appropriate unit of perceptual representation, evaluating the model output in terms of its utility, and incorporating cognitive constraints into the inference process. Our more cognitively plausible model shows a beneficial effect of cognitive constraints on segmentation performance. One interpretation of this effect is as a synergy between the naive theories of language structure that infants may have and the cognitive constraints that limit the fidelity of their inference processes, where less accurate inference approximations are better when the underlying assumptions about how words are generated are less accurate. More generally, these results highlight the utility of incorporating cognitive plausibility more fully into computational models of language acquisition. (shrink)

Pi-Chuan Chang, Michel Galley, and Christopher D. Manning Computer Science Department, Stanford University Stanford, CA 94305 pichuan,galley,[email protected]..

It has been well documented how language‐specific cues may be used for word segmentation. Here, we investigate what role a language‐independent phonological universal, the sonority sequencing principle (SSP), may also play. Participants were presented with an unsegmented speech stream with non‐English word onsets that juxtaposed adherence to the SSP with transitional probabilities. Participants favored using the SSP in assessing word‐hood, suggesting that the SSP represents a potentially powerful cue for word segmentation. To ensure the (...) SSP influenced the segmentation process (i.e., during learning), we presented two additional groups of participants with either (a) no exposure to the stimuli prior to testing or (b) the same stimuli with pauses marking word breaks. The SSP did not influence test performance in either case, suggesting that the SSP is important for word segmentation during the learning process itself. Moreover, the fact that SSP‐independent segmentation of the stimulus occurred (in the latter control condition) suggests that universals are best understood as biases rather than immutable constraints on learning. (shrink)

It has been well documented how language-specific cues may be used for word segmentation. Here, we investigate what role a language-independent phonological universal, the sonority sequencing principle (SSP), may also play. Participants were presented with an unsegmented speech stream with non-English word onsets that juxtaposed adherence to the SSP with transitional probabilities. Participants favored using the SSP in assessing word-hood, suggesting that the SSP represents a potentially powerful cue for word segmentation. To ensure the (...) SSP influenced the segmentation process (i.e., during learning), we presented two additional groups of participants with either (a) no exposure to the stimuli prior to testing or (b) the same stimuli with pauses marking word breaks. The SSP did not influence test performance in either case, suggesting that the SSP is important for word segmentation during the learning process itself. Moreover, the fact that SSP-independent segmentation of the stimulus occurred (in the latter control condition) suggests that universals are best understood as biases rather than immutable constraints on learning. (shrink)

In alphabetic writing systems (such as English), the spaces between words mark the word boundaries, and the basic unit of reading is distinguished during visual-level processing. The visual-level information of word boundaries facilitates reading. Chinese is an ideographic language whose text contains no intrinsic inter-word spaces as the marker of word boundaries. Previous studies have shown that the basic processing unit of Chinese reading is also a word. However, findings remain inconsistent regarding whether inserting spaces (...) between words in Chinese text promotes reading performance. Researchers have proposed that there may be a trade-off between format familiarity and the facilitation effect of inter-word spaces. In order to verify this, this study manipulated the format familiarity via reversing the Chinese reading direction from right to left to investigate this issue in Experiment 1 and Experiment 2. The purpose of Experiment 1 was to examine whether inter-word spaces facilitated Chinese reading in an unfamiliar format. Experiment 1 was conducted that 40 native Chinese undergraduates read Chinese sentences from right to left on four format conditions. The results showed faster reading speed and shorter total reading time for the inter-word spaced format. Based on this finding, Experiment 2 examined whether the facilitation effect of inter-word spaces would reduce or disappear after improving the format familiarity; this experiment was conducted that 40 native Chinese undergraduates who did not participate in Experiment 1 read Chinese sentences from right to left on four format conditions after ten-day reading training. There was no significant difference between the total reading time and reading speed in the inter-word spaced format and unspaced format, which suggests that the facilitation effect of inter-word spaces in Chinese reading changed smaller. The combined results of the two experiments suggest that there is indeed a trade-off between format familiarity and the facilitation of word segmentation, which supports the assumption of previous studies. (shrink)

A prerequisite for spoken language learning is segmenting continuous speech into words. Amongst many possible cues to identify word boundaries, listeners can use both transitional probabilities between syllables and various prosodic cues. However, the relative importance of these cues remains unclear, and previous experiments have not directly compared the effects of contrasting multiple prosodic cues. We used artificial language learning experiments, where native German speaking participants extracted meaningless trisyllabic “words” from a continuous speech stream, to evaluate these factors. We (...) compared a baseline condition to five test conditions, in which word-final syllables were either followed by a pause, lengthened, shortened, changed to a lower pitch, or changed to a higher pitch. To evaluate robustness and generality we used three tasks varying in difficulty. Overall, pauses and final lengthening were perceived as converging with the statistical cues and facilitated speech segmentation, with pauses helping most. Final-syllable shortening hindered baseline speech segmentation, indicating that when cues conflict, prosodic cues can override statistical cues. Surprisingly, pitch cues had little effect, suggesting that duration may be more relevant for speech segmentation than pitch in our study context. We discuss our findings with regard to the contribution to speech segmentation of language-universal boundary cues vs. language-specific stress patterns. (shrink)

When infants and toddlers are confronted with sequences of sounds, they are required to segment the sounds into meaningful units to achieve sufficient understanding. Rhythm has been regarded as a crucial cue for segmentation of speech sounds. Although previous intermodal methods indicated that infants and toddlers could detect differences in speech sounds based on stress-timed and syllable-timed units, these methods could not clearly indicate how infants and toddlers perform sound segmentation. Thus, the present study examined whether Japanese infants (...) and toddlers could segment word speech sounds comprising basic morae, on the basis of concurrent basic mora units within syllable units, using the new intermodal matching procedure. The results indicated that, regardless of their ages and linguistic abilities, Japanese infants and toddlers aged 6–25 months tended to segment Japanese words comprising basic morae sounds on the basis of concurrent basic mora units within syllable units. This implies that infants' and toddlers' use of syllable units for segmentation of speech sounds at an early age could be evident among many infants and toddlers learning various languages. Although this finding should be interpreted carefully, the present study demonstrated the utility of the new intermodal matching procedure for examining segmentation of speech sounds and word sounds by infants and toddlers, on the basis of specific rhythm units. (shrink)

In this paper, a novel approach for word extraction and character segmentation from the handwritten Bangla document images is reported. At first, a modified Run Length Smoothing Algorithm, called Spiral Run Length Smearing Algorithm, is applied for the extraction of words from the text lines of unconstrained handwritten Bangla document images. This technique has helped to overcome some of the drawbacks of standard horizontal and vertical RLSA techniques. SRLSA technique has been applied on the Bangla handwritten document image (...) database CMATERdb1.1.1 and the success rate of the word extraction is found to be 86.01%. In the second part of the work, we have presented a useful solution to the problem on how best word images of handwritten Bangla script can be segmented into constituent characters. Moreover, the technique can segment the words having discontinuity in Matra, a prominent feature of Bangla script. It also optimizes the trade-off between under/over segmentation as Matra region and segmentation points are estimated more precisely. As a result, better word segmentation accuracy is achieved with minimal data loss. Here, a success rate of 92.48% is observed on a dataset of 750 handwritten Bangla words which is 3.35% higher than that of our earlier techniques. (shrink)

Computational models of infant word‐finding typically operate over transcriptions of infant‐directed speech corpora. It is now possible to test models of word segmentation on speech materials, rather than transcriptions of speech. We propose that such modeling efforts be conducted over the speech of the experimental stimuli used in studies measuring infants' capacity for learning from spoken sentences. Correspondence with infant outcomes in such experiments is an appropriate benchmark for models of infants. We demonstrate such an analysis by (...) applying the DP‐Parser model of Algayres and colleagues to auditory stimuli used in infant psycholinguistic experiments by Pelucchi and colleagues. The DP‐Parser model takes speech as input, and creates multiple overlapping embeddings from each utterance. Prospective words are identified as clusters of similar embedded segments. This allows segmentation of each utterance into possible words, using a dynamic programming method that maximizes the frequency of constituent segments. We show that DP‐Parse mimics American English learners' performance in extracting words from Italian sentences, favoring the segmentation of words with high syllabic transitional probability. This kind of computational analysis over actual stimuli from infant experiments may be helpful in tuning future models to match human performance. (shrink)

Infants are sensitive to statistics in spoken language that aid word‐form segmentation and immediate mapping to referents. However, it is not clear whether this sensitivity influences the formation and retention of word‐referent mappings across a delay, two real‐world challenges that learners must overcome. We tested how the timing of referent training, relative to familiarization with transitional probabilities (TPs) in speech, impacts English‐learning 23‐month‐olds’ ability to form and retain word‐referent mappings. In Experiment 1, we tested infants’ ability (...) to retain TP information across a 10‐min delay and use it in the service of word learning. Infants successfully mapped high‐TP but not low‐TP words to referents. In Experiment 2, infants readily mapped the same words even when they were unfamiliar. In Experiment 3, high‐ and low‐TP word‐referent mappings were trained immediately after familiarization, and infants readily remembered these associations 10 min later. In sum, although 23‐month‐old infants do not need strong statistics to map word forms to referents immediately, or to remember those mappings across a delay, infants are nevertheless sensitive to these statistics in the speech stream, and they influence mapping after a delay. These findings suggest that, by 23 months of age, sensitivity to statistics in speech may impact infants’ language development by leading word forms with low coherence to be poorly mapped following even a short period of consolidation. (shrink)

This paper reconsiders the diphone-based word segmentation model of Cairns, Shillcock, Chater, and Levy (1997) and Hockema (2006), previously thought to be unlearnable. A statistically principled learning model is developed using Bayes’ theorem and reasonable assumptions about infants’ implicit knowledge. The ability to recover phrase-medial word boundaries is tested using phonetic corpora derived from spontaneous interactions with children and adults. The (unsupervised and semi-supervised) learning models are shown to exhibit several crucial properties. First, only a small amount (...) of language exposure is required to achieve the model’s ceiling performance, equivalent to between 1 day and 1 month of caregiver input. Second, the models are robust to variation, both in the free parameter and the input representation. Finally, both the learning and baseline models exhibit undersegmentation, argued to have significant ramifications for speech processing as a whole. (shrink)

Recent work asked if visual cues facilitate word segmentation in adults and infants. While adults showed better word segmentation when presented with a regular visual cue, infants did not. This difference was attributed to infants' lack of understanding that objects have labels. Alternatively, infants’ performance could reflect their difficulty with tracking and integrating multiple multimodal cues. We contrasted these two accounts by looking at the effect of visual cues on word segmentation in adults and (...) across childhood. We found that older children benefitted from the regular visual cues, but younger children, who already knew that objects have labels, did not. Knowing that objects have labels was not enough to use visual cues as an aid for segmentation. These findings show that the ability to integrate multimodal cues develops during childhood, and it is not yet adult-like in children. (shrink)

This study investigates a strategy based on predictability of consecutive sub-lexical units in learning to segment a continuous speech stream into lexical units using computational modeling and simulations. Lexical segmentation is one of the early challenges during language acquisition, and it has been studied extensively through psycholinguistic experiments as well as computational methods. However, despite strong empirical evidence, the explicit use of predictability of basic sub-lexical units in models of segmentation is underexplored. This paper presents an incremental computational (...) model of lexical segmentation for exploring the usefulness of predictability for lexical segmentation. We show that the predictability cue is a strong cue for segmentation. Contrary to earlier reports in the literature, the strategy yields state-of-the-art segmentation performance with an incremental computational model that uses only this particular cue in a cognitively plausible setting. The paper also reports an in-depth analysis of the model, investigating the conditions affecting the usefulness of the strategy. (shrink)

Infants have been described as 'statistical learners' capable of extracting structure (such as words) from patterned input (such as language). Here, we investigated whether prior knowledge influences how infants track transitional probabilities in word segmentation tasks. Are infants biased by prior experience when engaging in sequential statistical learning? In a laboratory simulation of learning across time, we exposed 9- and 10-month-old infants to a list of either disyllabic or trisyllabic nonsense words, followed by a pause-free speech stream composed (...) of a different set of disyllabic or trisyllabic nonsense words. Listening times revealed successful segmentation of words from fluent speech only when words were uniformly disyllabic or trisyllabic throughout both phases of the experiment. Hearing trisyllabic words during the pre-exposure phase derailed infants' abilities to segment speech into disyllabic words, and vice versa. We conclude that prior knowledge about word length equips infants with perceptual expectations that facilitate efficient processing of subsequent language input. (shrink)

In this paper, we study the performance of baseline hidden Markov model (HMM) for segmentation of speech signals. It is applied on single-speaker segmentation task, using Hindi speech database. The automatic phoneme segmentation framework evolved imitates the human phoneme segmentation process. A set of 44 Hindi phonemes were chosen for the segmentation experiment, wherein we used continuous density hidden Markov model (CDHMM) with a mixture of Gaussian distribution. The left-to-right topology with no skip states has (...) been selected as it is effective in speech recognition due to its consistency with the natural way of articulating the spoken words. This system accepts speech utterances along with their orthographic “transcriptions” and generates segmentation information of the speech. This corpus was used to develop context-independent hidden Markov models (HMMs) for each of the Hindi phonemes. The system was trained using numerous sentences that are relevant to provide information to the passengers of the Metro Rail. The system was validated against a few manually segmented speech utterances. The evaluation of the experiments shows that the best performance is obtained by using a combination of two Gaussians mixtures and five HMM states. A category-wise phoneme error analysis has been performed, and the performance of the phonetic segmentation has been reported. The modeling of HMMs has been implemented using Microsoft Visual Studio 2005 (C++), and the system is designed to work on Windows operating system. The goal of this study is automatic segmentation of speech at phonetic level. (shrink)

Infants are sensitive to statistical regularities (i.e., transitional probabilities, or TPs) relevant to segmenting words in fluent speech. However, there is debate about whether tracking TPs results in representations of possible words. Infants show preferential learning of sequences with high TPs (HTPs) as object labels relative to those with low TPs (LTPs). Such findings could mean that only the HTP sequences have a word‐like status, and they are more readily mapped to a referent for that reason. But these findings (...) could also suggest that HTP sequences are easier to encode, just like any other predictable sequence. Here we aimed to distinguish between these explanations. To do so, we built on findings that infants become resistant to learning labels that are not typical of their native language as they approach 2 years of age and add words to their lexicons. If tracking TPs in speech results in identifying candidate words, at this age TPs may have reduced power to confer lexical status when they yield a unit that is very dissimilar to word forms that are typical of infants’ native language. Indeed, we found that at 20 months, English‐learning infants with relatively small vocabularies learned HTP Italian words (but not LTP words) as object labels, while infants with larger vocabularies resisted learning HTP Italian words. These findings suggest that the HTP sequences may be represented as candidate words, and more broadly, that TP statistics are relevant to word learning. (shrink)

With two eye movements tracking experiments, present study sought to establish whether predictability of target words facilitate characters processing in Chinese reading. Target Words embedded in sentences in both experiments were composed by 2 Chinese characters. Predictability of target words were manipulated in both experiments, beyond that, frequency of targets' first characters were manipulated in Experiment 1, frequency of targets' second characters were manipulated in Experiment 2. Pervasive predictability effects were observed on almost all of eye movement measures in both (...) experiments. It also yield reliable characters frequency effect in Experiment 1 & 2, on some measures. None reliable interaction effects of predictability and first character frequency were observed in Experiment 1, but predictability and second character frequency interactively impacted on some measures of eye movement. Theoretical implications of the data are discussed. (shrink)

This paper presents an implemented computational model of word acquisition which learns directly from raw multimodal sensory input. Set in an information theoretic framework, the model acquires a lexicon by finding and statistically modeling consistent cross‐modal structure. The model has been implemented in a system using novel speech processing, computer vision, and machine learning algorithms. In evaluations the model successfully performed speech segmentation, word discovery and visual categorization from spontaneous infant‐directed speech paired with video images of single (...) objects. These results demonstrate the possibility of using state‐of‐the‐art techniques from sensory pattern recognition and machine learning to implement cognitive models which can process raw sensor data without the need for human transcription or labeling. (shrink)

Vowel phonemes are a part of any acoustic speech signal. Vowel sounds occur in speech more frequently and with higher energy. Therefore, vowel phoneme can be used to extract different amounts of speaker discriminative information in situations where acoustic information is noise corrupted. This article presents an approach to identify a speaker using the vowel sound segmented out from words spoken by the speaker. The work uses a combined self-organizing map - and probabilistic neural network -based approach to segment the (...) vowel phoneme. The segmented vowel is later used to identify the speaker of the word by matching the patterns with a learning vector quantization -based code book. The LVQ code book is prepared by taking features of clean vowel phonemes uttered by the male and female speakers to be identified. The proposed work formulates a framework for the design of a speaker-recognition model of the Assamese language, which is spoken by ∼3 million people in the Northeast Indian state of Assam. The experimental results show that the segmentation success rates obtained using a SOM-based technique provides an increase of at least 7% compared with the discrete wavelet transform-based technique. This increase contributes to the improvement in overall performance of speaker identification by ∼3% compared with earlier related works. (shrink)

From an early age, exposure to a spoken language has allowed us to implicitly capture the structure underlying the succession of speech sounds in that language and to segment it into meaningful units. Statistical learning, the ability to pick up patterns in the sensory environment without intention or reinforcement, is thus assumed to play a central role in the acquisition of the rule-governed aspects of language, including the discovery of word boundaries in the continuous acoustic stream. Although extensive evidence (...) has been gathered from artificial languages experiments showing that children and adults are able to track the regularities embedded in the auditory input, as the probability of one syllable to follow another syllable in the speech stream, the developmental trajectory of this ability remains controversial. In this work, we have collected Event-Related Potentials while 5-year-old children and young adults were exposed to a speech stream made of the repetition of eight three-syllable nonsense words presenting different levels of predictability to mimic closely what occurs in natural languages and to get new insights into the changes that the mechanisms underlying auditory statistical learning might undergo through the development. The participants performed the aSL task first under implicit and, subsequently, under explicit conditions to further analyze if children take advantage of previous knowledge of the to-be-learned regularities to enhance SL, as observed with the adult participants. These findings would also contribute to extend our knowledge of the mechanisms available to assist SL at each developmental stage. Although behavioral signs of learning, even under explicit conditions, were only observed for the adult participants, ERP data showed evidence of online segmentation in the brain in both groups, as indexed by modulations in the N100 and N400 components. A detailed analysis of the neural data suggests, however, that adults and children rely on different mechanisms to assist the extraction of word-like units from the continuous speech stream, hence supporting the view that SL with auditory linguistic materials changes through development. (shrink)