Phoneme Data Research Articles

A RTL Implementation of Heterogeneous Machine Learning Network for French Computer Assisted Pronunciation Training

Computer-assisted pronunciation training (CAPT) is a helpful method for self-directed or long-distance foreign language learning. It greatly benefits from the progress, and of acoustic signal processing and artificial intelligence techniques. However, in real-life applications, embedded solutions are usually desired. This paper conceives a register-transfer level (RTL) core to facilitate the pronunciation diagnostic tasks by suppressing the mulitcollinearity of the speech waveforms. A recently proposed heterogeneous machine learning framework is selected as the French phoneme pronunciation diagnostic algorithm. This RTL core is implemented and optimized within a very-high-level synthesis method for fast prototyping. An original French phoneme data set containing 4830 samples is used for the evaluation experiments. The experiment results demonstrate that the proposed implementation reduces the diagnostic error rate by 0.79–1.33% compared to the state-of-the-art and achieves a speedup of 10.89× relative to its CPU implementation at the same abstract level of programming languages.

Verbal fluency as a quick and simple tool to help in deciding when to refer patients with a possible brain tumour

BackgroundPatients with brain tumours often present with non-specific symptoms. Correctly identifying who to prioritise for urgent brain imaging is challenging. Brain tumours are amongst the commonest cancers diagnosed as an emergency presentation. A verbal fluency task (VFT) is a rapid triage test affected by disorders of executive function, language and processing speed. We tested whether a VFT could support identification of patients with a brain tumour.MethodsThis proof-of-concept study examined whether a VFT can help differentiate patients with a brain tumour from those with similar symptoms (i.e. headache) without a brain tumour. Two patient populations were recruited, (a) patients with known brain tumour, and (b) patients with headache referred for Direct-Access Computed-Tomography (DACT) from primary care with a suspicion of a brain tumour. Semantic and phonemic verbal fluency data were collected prospectively.Results180 brain tumour patients and 90 DACT patients were recruited. Semantic verbal fluency score was significantly worse for patients with a brain tumour than those without (P < 0.001), whether comparing patients with headache, or patients without headache. Phonemic fluency showed a similar but weaker difference. Raw and incidence-weighted positive and negative predictive values were calculated.ConclusionWe have demonstrated the potential role of adding semantic VFT score performance into clinical decision making to support triage of patients for urgent brain imaging. A relatively small improvement in the true positive rate in patients referred for DACT has the potential to increase the timeliness and efficiency of diagnosis and improve patient outcomes.

Impacts of multicollinearity on CAPT modalities: An heterogeneous machine learning framework for computer-assisted French phoneme pronunciation training.

Phoneme pronunciations are usually considered as basic skills for learning a foreign language. Practicing the pronunciations in a computer-assisted way is helpful in a self-directed or long-distance learning environment. Recent researches indicate that machine learning is a promising method to build high-performance computer-assisted pronunciation training modalities. Many data-driven classifying models, such as support vector machines, back-propagation networks, deep neural networks and convolutional neural networks, are increasingly widely used for it. Yet, the acoustic waveforms of phoneme are essentially modulated from the base vibrations of vocal cords, and this fact somehow makes the predictors collinear, distorting the classifying models. A commonly-used solution to address this issue is to suppressing the collinearity of predictors via partial least square regressing algorithm. It allows to obtain high-quality predictor weighting results via predictor relationship analysis. However, as a linear regressor, the classifiers of this type possess very simple topology structures, constraining the universality of the regressors. For this issue, this paper presents an heterogeneous phoneme recognition framework which can further benefit the phoneme pronunciation diagnostic tasks by combining the partial least square with support vector machines. A French phoneme data set containing 4830 samples is established for the evaluation experiments. The experiments of this paper demonstrates that the new method improves the accuracy performance of the phoneme classifiers by 0.21 − 8.47% comparing to state-of-the-arts with different data training data density.

KORESPONDENSI FONEMIS BAHASA OGAN DAN BAHASA BANGKA

The objects of this research are Organese and Bangkanese language. The objectives of this research are to describe phonemic correspondence of Organese and Bangkanese language, and to compare the phonemic correspondence formula of both languages. The data of this research are phonemes from both languages which was gathered using dictionary study method with taking-note technique. The phoneme data searched were phonemes in which they have correspondence value. There are several findings in this research, they are: (1) proto-phoneme */u/ retained in Bangkanese, while in Organese is reflected as phoneme /o/. The reflection occurs in closed-ultima positions, if followed by bilabial plosive-stop /m/, nasal dorso-velar /ŋ/, glottal slide /h/. dorso-velar plosive-stop /k/, and apicodental plosive stop /t/; (2) proto-phoneme */a/ retained in Organese, while in Bangkanese is reflected as /Ə/. The reflection occurs in closed-ultima positions, if followed by bilabial plosive-stop /p/, glottal slide /h/, apico-lateral /r/, nasal dorso-velar /ŋ/, bilabial plosive-stop /m/, apicodental plosive-stop /t/, lateral /l/, nasal apico-alveolar /n/, and glottal plosive-stop /Ɂ/; and (3) proto-phoneme */h/ retained in Bangkanese, while in Organese is either reflected as /ø/ or being eliminated. The particular reflection occurs in penultima positions, if followed by vowels /a/, /i/, and /u/.

Recursive non-parametric kernel classification rule estimation for independent functional data

In this paper we propose an automatic selection of the bandwidth of the recursive non-parametric estimation of the kernel classification rule function defined by the stochastic approximation algorithm, when the explanatory data are curves and the response is categorical. We established a central limit theorem for our proposed recursive estimators, the proposed recursive estimators will be very competitive to the non-recursive one in terms of estimation error but much better in terms of computational costs. The proposed estimators are used first on simulated waveform curves and then on real phoneme data.

Fonem Deret Konsonan dalam Bahasa Indonesia

The existence of consonant series phonemes in Indonesian phonological learning is important, but strangely it has not received special attention, so it is interesting to be the object of study. Based on existing data, if we find a phoneme study of consonant series from several references, it turns out that the discussion is only a part / sub-section of the subject matter of the study. The application of theory is based on the linguistic theory of phonological branches, phonemic subcabang, because the study focuses on the phoneme of the consonant series itself and its ability to form consonant sequences. The implementation of the study starts at three strategic stages as in linguistic research in general, namely: 1. the stage of providing data, 2. the stage of classification and analysis of data, and 3. the stage of preparing the report. The phoneme data of consonant series is more secondary, so the existence of references is a data source. The results of the consonant phoneme data analysis almost apply to all consonant phonemes in Indonesian, including loan / absorption (foreign) phonemes. The basis for determining the consonant phoneme starts with: 1. a consonant series is a consonant series of two / or more phonemes in one word, but in different syllables; 2. consonant phoneme data departs from the kinds of vocabulary in Indonesian; and 3. that the question of syllable decapitation is based on decapitation of the word form (morphological), not on sound (phonological) elements. The types of vocabulary used as data include: 1) original Indonesian vocabulary; 2) loan / absorption vocabulary, especially from the process "received by adjustment"; and 3) the "new form" vocabulary.

Inference on Covariance Operators via Concentration Inequalities: k-sample Tests, Classification, and Clustering via Rademacher Complexities

We propose a novel approach to the analysis of covariance operators making use of concentration inequalities. First, non-asymptotic confidence sets are constructed for such operators. Then, subsequent applications including a k sample test for equality of covariance, a functional data classifier, and an expectation-maximization style clustering algorithm are derived and tested on both simulated and phoneme data.

Rejection of a serial founder effects model of genetic and linguistic coevolution

Recent genetic studies attribute the negative correlation between population genetic diversity and distance from Africa to a serial founder effects (SFE) evolutionary process. A recent linguistic study concluded that a similar decay in phoneme inventories in human languages was also the product of the SFE process. However, the SFE process makes additional predictions for patterns of neutral genetic diversity, both within and between groups, that have not yet been tested on phonemic data. In this study, we describe these predictions and test them on linguistic and genetic samples. The linguistic sample consists of 725 widespread languages, which together contain 908 distinct phonemes. The genetic sample consists of 614 autosomal microsatellite loci in 100 widespread populations. All aspects of the genetic pattern are consistent with the predictions of SFE. In contrast, most of the predictions of SFE are violated for the phonemic data. We show that phoneme inventories provide information about recent contacts between languages. However, because phonemes change rapidly, they cannot provide information about more ancient evolutionary processes.

바타챠랴 거리 측정 기법을 사용한 가우시안 모델 기반 음소 인식 향상

기존의 어휘 인식에서는 일반적인 벡터 값을 데이터베이스를 이용하여 구하므로 탐색 중에 형성되는 음소를 처리하지 못하는 문제점을 제공하며, 음소 데이터에 대한 모델을 구성할 수 없는 단점으로 인하여 가우시안 모텔의 정확성을 확보하지 못하게 된다. 따라서 본 논문에서는 음소가 갖는 특징을 기반으로 바타챠랴 거리 측정법을 이용하여 정확한 음소로 인식할 수 있도록 유도하였으며 유사 음소 인식과 오인식 오류를 최소화하여 인식률을 향상시켰다. 연속 확률 분포의 공유로부터 가우시안 모델 최적화를 실험한 결과 향상된 신뢰도로 인해 높은 인식 성능을 확인하였으며, 본 논문에서 제안한 바타챠랴 거리 측정법을 이용하여 실험한 결과 기존의 방법들에 비하여 평균 1.9%의 성능 향상을 나타내었으며 신뢰성을 바탕으로 인식율에서 평균 2.9%의 성능 향상을 나타내었다. Previous existing vocabulary recognition programs calculate general vector values from a database, so they can not process phonemes that form during a search. And because they can not create a model for phoneme data, the accuracy of the Gaussian model can not secure. Therefore, in this paper, we recommend use of the Bhattacharyya distance measurement method based on the features of the phoneme-thus allowing us to improve the recognition rate by picking up accurate phonemes and minimizing recognition of similar and erroneous phonemes. We test the Gaussian model optimization through share continuous probability distribution, and we confirm the heighten recognition rate. The Bhattacharyya distance measurement method suggest in this paper reflect an average 1.9% improvement in performance compare to previous methods, and it has average 2.9% improvement based on reliability in recognition rate.

미등록어 거절 알고리즘에서 가우시안 모델 최적화를 이용한 신뢰도 정규화 향상

어휘 인식에서는 인식 학습 시 나타나지 않는 미 출현 트라이 폰이 존재하며, 이들 시스템에서는 모델 파라미터들의 초기 추정치를 생성하지 못하고 음소 데이터에 대한 모델을 구성할 수 없는 단점으로 인하여 가우시안 모델의 정확성을 확보하지 못하게 된다. 이를 개선하기 위하여 확률 분포를 이용한 모델 파라미터의 가우시안 모델 최적화 방법을 제안한다. 확률 분포의 가우시안 모델을 최적화하여 가우시안 모델의 정확성을 제공하고, 음소 단위로 데이터의 탐색을 지원하여 신뢰도가 향상되었다. 제안된 방법의 성능 평가를 위하여 실제 다양한 미등록어가 관측될 수 있는 대상으로 실험을 수행하였으며 본 연구에서 제안한 정규화 신뢰도를 이용한 미등록어 거절 알고리즘이 기존의 방법들에 비하여 평균 1.7%의 성능향상을 나타내었다. In vocabulary recognition has unseen tri-phone appeared when recognition training. This system has not been created beginning estimation figure of model parameter. It's bad points could not be created that model for phoneme data. Therefore it's could not be secured accuracy of Gaussian model. To improve suggested Gaussian model to optimized method of model parameter using probability distribution. To improved of confidence that Gaussian model to optimized of probability distribution to offer by accuracy and to support searching of phoneme data. This paper suggested system performance comparison as a result of recognition improve represent 1.7% by out-of vocabulary rejection algorithm using normalization confidence.

음소 유사율 오류 보정을 이용한 어휘 인식 후처리 시스템

어휘 인식 시스템에서 인식률 저하의 요인으로는 유사한 음소 인식과 부정확한 어휘 제공으로 인해 오인식 오류가 존재한다. 부정확한 어휘의 입력으로 특징을 추출하여 인식할 경우 오인식의 결과가 나타나거나 유사한 음소로 인식되며 특징 추출이 제대로 이루어지지 않으면 음소 인식 시 유사한 음소로 인식하게 된다. 따라서 본 논문에서는 음소가 갖는 특징을 기반으로 음소 유사율을 이용한 어휘 인식 후처리에서의 오류 보정 후처리 시스템을 제안하였다. 음소 유사율은 모노폰으로 훈련시킨 훈련 데이터를 각각의 음소에 MFCC와 LPC 특징 추출 방법을 이용하여 구하였다. 유사한 음소는 정확한 음소로 인식할 수 있도록 유도하여 부정확한 어휘 제공으로 인하여 오인식되는 오류를 최소화하였다. 음소 유사율과 신뢰도를 이용하여 오류 보정율을 구하였으며, 어휘 인식 과정에서 오류로 판명된 어휘에 대하여 오류 보정을 수행하였다. 에러패턴 학습을 이용한 시스템과 의미기반을 이용한 시스템에 비해 시스템 성능 평가 결과 MFCC와 LPC는 각각 7.5%와 5.3%의 인식 향상률을 보였다. In vocabulary recognition system has reduce recognition rate unrecognized error cause of similar phoneme recognition and due to provided inaccurate vocabulary. Input of inaccurate vocabulary by feature extraction case of recognition by appear result of unrecognized or similar phoneme recognized. Also can't feature extraction properly when phoneme recognition is similar phoneme recognition. In this paper propose vocabulary recognition post-process error correction system using phoneme likelihood based on phoneme feature. Phoneme likelihood is monophone training phoneme data by find out using MFCC and LPC feature extraction method. Similar phoneme is induced able to recognition of accurate phoneme due to inaccurate vocabulary provided unrecognized reduced error rate. Find out error correction using phoneme likelihood and confidence when vocabulary recognition perform error correction for error proved vocabulary. System performance comparison as a result of recognition improve represent MFCC 7.5%, LPC 5.3% by system using error pattern and system using semantic.

형상 형성 제어를 이용한 어휘인식 공유 모델의 가우시안 최적화

연속 어휘 인식 확률 분포의 공유 방법에서는 사용될 모델 파라미터들의 초기 추정치를 생성하기 위한 각 문맥들에 대한 음소 데이터가 반드시 필요하지만 이들 음소 데이터에 대한 모델을 구성할 수 없는 단점으로 가우시안 모델의 정확성을 확보하지 못한다는 단점이 있다. 이를 개선하기 위하여 확률 분포의 혼합 가우시안 모델을 최적화하고, 음소 단위로 데이터를 탐색을 지원하는 형상 형성 시스템을 제안한다. 본 논문의 형상 형성 시스템은 확장 facet 분류를 이용하여 사용자에게 음소 단위의 형상 형성 정보를 제공하므로 가우시안 모델의 정확성을 제공한다. 본 논문에서 제안한 시스템을 적용한 결과 시스템 성능에서 어휘 종속 인식률은 98.31%, 어휘 독립 인식률은 97.63%의 인식률을 나타내었다. In continuous vocabulary recognition system by probability distribution of clustering method has used model parameters of an advance estimate to generated each contexts for phoneme data surely needed but it has it's bad points of gaussian model the accuracy unsecure of composed model for phoneme data. To improve suggested probability distribution mixed gaussian model to optimized that phoneme data search supported configuration thread system. This paper of configuration thread system has used extension facet classification user phoneme configuration thread information offered gaussian model the accuracy secure. System performance as a result of represent vocabulary dependence recognition rate of 98.31%, vocabulary independence recognition rate of 97.63%.

R-POPTVR: A Novel Reinforcement-Based POPTVR Fuzzy Neural Network for Pattern Classification

In general, a fuzzy neural network (FNN) is characterized by its learning algorithm and its linguistic knowledge representation. However, it does not necessarily interact with its environment when the training data is assumed to be an accurate description of the environment under consideration. In interactive problems, it would be more appropriate for an agent to learn from its own experience through interactions with the environment, i.e., reinforcement learning. In this paper, three clustering algorithms are developed based on the reinforcement learning paradigm. This allows a more accurate description of the clusters as the clustering process is influenced by the reinforcement signal. They are the REINFORCE clustering technique I (RCT-I), the REINFORCE clustering technique II (RCT-II), and the episodic REINFORCE clustering technique (ERCT). The integrations of the RCT-I, the RCT-II, and the ERCT within the pseudo-outer product truth value restriction (POPTVR), which is a fuzzy neural network integrated with the truth restriction value (TVR) inference scheme in its five layered feedforward neural network, form the RPOPTVR-I, the RPOPTVR-II, and the ERPOPTVR, respectively. The Iris, Phoneme, and Spiral data sets are used for benchmarking. For both Iris and Phoneme data, the RPOPTVR is able to yield better classification results which are higher than the original POPTVR and the modified POPTVR over the three test trials. For the Spiral data set, the RPOPTVR-II is able to outperform the others by at least a margin of 5.8% over multiple test trials. The three reinforcement-based clustering techniques applied to the POPTVR network are able to exhibit the trial-and-error search characteristic that yields higher qualitative performance.

Speech synthesizing apparatus and method, and storage medium therefor

A speech synthesizing apparatus for synthesizing a speech waveform stores speech data, which is obtained by adding attribute information onto phoneme data, in a database. In accordance with prescribed retrieval conditions, a phoneme retrieval unit retrieves phoneme data from the speech data that has been stored in the database and retains the retrieved results in a retrieved-result storage area. A processing unit for assigning a power penalty and a processing unit for assigning a phoneme-duration penalty assign the penalties, on the basis of power and phoneme duration constituting the attribute information, to a set of phoneme data stored in the retrieved-result storage area. A processing unit for determining typical phoneme data performs sorting on the basis of the assigned penalties and, based upon the stored results, selects phoneme data to be employed in the synthesis of a speech waveform.

Speech synthesis apparatus, control method therefor and computer-readable memory

A second phoneme is generated in consideration of a phonemic context with respect to a first phoneme as a search target. Phonemic piece data corresponding to the second phoneme is searched out from a database. A third phoneme is generated by changing the phonemic context on the basis of the search result, and phonemic piece data corresponding to the third phoneme is re-searched out from the database. The search or re-search result is registered in a table in correspondence with the second or third phoneme.

Computer apparatus for text-to-speech synthesizer dictionary reduction

A computerized apparatus for reducing the size of a dictionary used in a text-to-speech synthesis system are provided. In an initial phase, the method and apparatus determine if entries in the dictionary, each containing a grapheme string and a corresponding phoneme string, can be fully matched by using at least one rule set used to synthesize words to phonemic data. If the entry can be fully matched using rule processing alone, the entry is indicated to be deleted from the dictionary. In a second phase, the method and apparatus determine if the entry, considered as a root word entry, is required in the dictionary in order to support phoneme synthesis of other entries containing the root word entry, and if so, the root word entry is indicated to be saved in the dictionary. If the other entries containing the root word entry can have correct phonemic data generated from a combination of the root word entries phonemic data and phonemes generated from rule set processing, then the other entries are indicated to be deleted from the dictionary. After all words have been processed by phase one and/or phase two, the entries indicated to be saved are aggregated to form a reduced dictionary.

A novel approach to the derivation of fuzzy membership functions using the Falcon-MART architecture

A fuzzy neural network, Falcon-MART, is proposed in this paper. This is a modification of the original Falcon-ART architecture. Both Falcon-ART and Falcon-MART are fuzzy neural networks that can be used as fuzzy controllers or applied to areas such as forgery detection, pattern recognition and data analysis. They constitute a group of hybrid systems that incorporate fuzzy logic into neural networks. In this way, the structure of these hybrid networks become transparent as high level IF-THEN human-like reasoning is used to interpret the network connections. In addition, the hybrid networks automatically derive the fuzzy rules (knowledge base) of the problem domain using neural network techniques and hence avoid the pitfalls of traditional fuzzy systems. The main problem in designing a fuzzy neural network is how to formulate the fuzzy rule base. Most proposed fuzzy neural networks in the literature could be classified into two categories. The first group assumes the existence of a preliminary rule base and uses neural techniques to tune the parameters to obtain the final set of fuzzy rules. The second group assumes no knowledge of any fuzzy rules and performs a cluster analysis on the numerical training data before formulating the rules from the computed clusters. Falcon-ART attempts to overcome the constraints faced by these two groups of fuzzy neural networks by using the fuzzy ART technique to partition the training data set. However, there are several shortcomings in the Falcon-ART network. They are: 1. Poor network performances when the classes of input data are closely similar to each other; 2. Weak resistance to noisy/spurious training data; 3. Termination of network training process depends heavily on a preset error parameter; and 4. Learning efficiency may deteriorate as a result of using complementary coded training data. Falcon-MART has been developed to address these shortcomings. To evaluate the effectiveness of Falcon-MART, three different sets of experiments are conducted. The first experiment demonstrates the efficiency of Falcon-MART over Falcon-ART using the Fisher's Iris data set. The second experiment evaluates the modeling capability of Falcon-MART against the classical multi-layered perceptron (MLP) network using a set of traffic flow data. The last experiment uses a set of phoneme data to demonstrate the clustering ability of Falcon-MART against the traditional K-nearest-neighbor ( K-NN) classifier. The results obtained are encouraging.

Reference pattern training system and speech recognition system using the same

A training unit selector 13 takes out and compares phoneme data about recognition units stored in a recognition word dictionary 10 and a training word dictionary 11, and selects a recognition unit from the recognition unit training words stored in the training word dictionary 11, the selected recognition unit resembling each selected recognition unit in a subject recognition word stored in the recognition word dictionary 10 in phoneme context, the selected recognition unit being sent to a recognition unit reference pattern generator 14. The recognition unit reference pattern generator 14 utilizes a recognition unit, which is best in accord in phoneme context among training data stored in a training data memory 12, concerning a recognition unit selected by the training unit selector 13.

A Connectionist Model for Bootstrap Learning of Syllabic Structure

We report on a series of experiments with simple recurrent networks (SRNs) solving phoneme prediction in continuous phonemic data. The purpose of the experiments is to investigate whether the network output could function as a source for syllable boundary detection. We show that this is possible, using a generalisation of the network resembling the linguistic sonority principle. We argue that the primary generalisation of the network, that is, the fact that sonority varies in a hat-shaped way across phonemic strings, ending and starting at syllable boundaries, is an indication that sonority might be a major cue in discovering the essential building bricks of language when confronted with unsegmented running speech. The segment which is most directly related to sonority patterns, the syllable, has received considerable attention in psycholinguistics as being an element of natural language that is easily grasped by language learners. The phoneme prediction network presents a simulation of the necessary bootstrap to arrive at the discovery of syllabic segmentation in unsegmented speech, which can be used as a basis for the segmentation of larger structures like words.

Method and apparatus for producing audio-visual synthetic speech

A method and apparatus provide a video image of facial features synchronized with synthetic speech. Text input is transformed into a string of phonemes and timing data, which are transmitted to an image generation unit. At the same time, a string of synthetic speech samples is transmitted to an audio server. The audio server produces signals for an audio speaker, causing the audio signals to be continuously audibilized; additionally, the audio server initializes a timer. The image generation unit reads the timing data from the timer and, by consulting the phoneme and timing data, determines the position of the phoneme currently being audibilized. The image generation unit then calculates the facial configuration corresponding to the position in the string of phonemes, calculates the facial configuration, and causes the facial configuration to be displayed on a video device.