Constructing accurate and robust HMM/GMM models for an Arabic speech recognition system

Conventional Hidden Markov Model (HMM) based Automatic Speech Recognition (ASR) systems generally utilize cepstral features as acoustic observation and phonemes as basic linguistic units. Some of the most powerful features currently used in ASR systems are Mel-Frequency Cepstral Coefficients (MFCCs). Speech recognition is inherently complicated due to the variability in the speech signal which includes within- and across-speaker variability. This leads to several kinds of mismatch between acoustic features and acoustic models and hence degrades the system performance. The sensitivity of MFCCs to speech signal variability motivates many researchers to investigate the use of a new set of speech feature parameters in order to make the acoustic models more robust to this variability and thus improve the system performance. The combination of diverse acoustic feature sets has great potential to enhance the performance of ASR systems. This paper is a part of ongoing research efforts aspiring to build an accurate Arabic ASR system for teaching and learning purposes. It addresses the integration of complementary features into standard HMMs for the purpose to make them more robust and thus improve their recognition accuracies. The complementary features which have been investigated in this work are voiced formants and Pitch in combination with conventional MFCC features. A series of experimentations under various combination strategies were performed to determine which of these integrated features can significantly improve systems performance. The Cambridge HTK tools were used as a development environment of the system and experimental results showed that the error rate was successfully decreased, the achieved results seem very promising, even without using language models.     

Domain-independent feature extraction for multi-classification using multi-objective genetic programming

We propose three model-free feature extraction approaches for solving the multiple class classification problem; we use multi-objective genetic programming (MOGP) to derive (near-)optimal feature extraction stages as a precursor to classification with a simple and fast-to-train classifier. Statistically-founded comparisons are made between our three proposed approaches and seven conventional classifiers over seven datasets from the UCI Machine Learning database. We also make comparisons with other reported evolutionary computation techniques. On almost all the benchmark datasets, the MOGP approaches give better or identical performance to the best of the conventional methods. Of our proposed MOGP-based algorithms, we conclude that hierarchical feature extraction performs best on multi-classification problems.     

Speaker recognition utilizing distributed DCT-II based Mel frequency cepstral coefficients and fuzzy vector quantization

In this paper, a new and novel Automatic Speaker Recognition (ASR) system is presented. The new ASR system includes novel feature extraction and vector classification steps utilizing distributed Discrete Cosine Transform (DCT-II) based Mel Frequency Cepstral Coefficients (MFCC) and Fuzzy Vector Quantization (FVQ). The ASR algorithm utilizes an approach based on MFCC to identify dynamic features that are used for Speaker Recognition (SR). A series of experiments were performed utilizing three different feature extraction methods: (1) conventional MFCC; (2) Delta-Delta MFCC (DDMFCC); and (3) DCT-II based DDMFCC. The experiments were then expanded to include four classifiers: (1) FVQ; (2) K-means Vector Quantization (VQ); (3) Linde, Buzo and Gray VQ; and (4) Gaussian Mixed Model (GMM). The combination of DCT-II based MFCC, DMFCC and DDMFCC with FVQ was found to have the lowest Equal Error Rate for the VQ based classifiers. The results found were an improvement over previously reported non-GMM methods and approached the results achieved for the computationally expensive GMM based method. Speaker verification tests carried out highlighted the overall performance improvement for the new ASR system. The National Institute of Standards and Technology Speaker Recognition Evaluation corpora was used to provide speaker source data for the experiments.     

Exploring the Use of Speech Features and Their Corresponding Distribution Characteristics for Robust Speech Recognition

The performance of current automatic speech recognition (ASR) systems often deteriorates radically when the input speech is corrupted by various kinds of noise sources. Several methods have been proposed to improve ASR robustness over the last few decades. The related literature can be generally classified into two categories according to whether the methods are directly based on the feature domain or consider some specific statistical feature characteristics. In this paper, we present a polynomial regression approach that has the merit of directly characterizing the relationship between speech features and their corresponding distribution characteristics to compensate for noise interference. The proposed approach and a variant were thoroughly investigated and compared with a few existing noise robustness approaches. All experiments were conducted using the Aurora-2 database and task. The results show that our approaches achieve considerable word error rate reductions over the baseline system and are comparable to most of the conventional robustness approaches discussed in this paper.     

Acoustic modeling problem for automatic speech recognition system: conventional methods (Part I)

In automatic speech recognition (ASR) systems, the speech signal is captured and parameterized at front end and evaluated at back end using the statistical framework of hidden Markov model (HMM). The performance of these systems depend critically on both the type of models used and the methods adopted for signal analysis. Researchers have proposed a variety of modifications and extensions for HMM based acoustic models to overcome their limitations. In this review, we summarize most of the research work related to HMM-ASR which has been carried out during the last three decades. We present all these approaches under three categories, namely conventional methods, refinements and advancements of HMM. The review is presented in two parts (papers): (i) An overview of conventional methods for acoustic phonetic modeling, (ii) Refinements and advancements of acoustic models. Part I explores the architecture and working of the standard HMM with its limitations. It also covers different modeling units, language models and decoders. Part II presents a review on the advances and refinements of the conventional HMM techniques along with the current challenges and performance issues related to ASR.     

Deep Learning with Optimal Hierarchical Spiking Neural Network for Medical Image Classification

Medical image classification becomes a vital part of the design of computer aided diagnosis (CAD) models. The conventional CAD models are majorly dependent upon the shapes, colors, and/or textures that are problem oriented and exhibited complementary in medical images. The recently developed deep learning (DL) approaches pave an efficient method of constructing dedicated models for classification problems. But the maximum resolution of medical images and small datasets, DL models are facing the issues of increased computation cost. In this aspect, this paper presents a deep convolutional neural network with hierarchical spiking neural network (DCNN-HSNN) for medical image classification. The proposed DCNN-HSNN technique aims to detect and classify the existence of diseases using medical images. In addition, region growing segmentation technique is involved to determine the infected regions in the medical image. Moreover, NADAM optimizer with DCNN based Capsule Network (CapsNet) approach is used for feature extraction and derived a collection of feature vectors. Furthermore, the shark smell optimization algorithm (SSA) based HSNN approach is utilized for classification process. In order to validate the better performance of the DCNN-HSNN technique, a wide range of simulations take place against HIS2828 and ISIC2017 datasets. The experimental results highlighted the effectiveness of the DCNN-HSNN technique over the recent techniques interms of different measures. Please type your abstract here.     

Automatic Speaker Recognition from Speech Signals Using Self Organizing Feature Map and Hybrid Neural Network

This paper explains a new hybrid method for Automatic Speaker Recognition using speech signals based on the Artificial Neural Network (ANN). ASR performance characteristics is regarded as the foremost challenge and necessitated to be improved. This research work mainly focusses on resolving the ASR problems as well as to improve the accuracy of the prediction of a speaker.. Mel Frequency Cepstral Coefficient (MFCC) is greatly exploited for signal feature extraction.The input samples are created using these extracted features and its dimensions have been reduced using Self Organizing Feature Map (SOFM). Finally, using the reduced input samples, recognition is performed using Multilayer Perceptron (MLP) with Bayesian Regularization.. The training of the network has been accomplished and verified by means of real speech datasets from the Multivariability speaker recognition database for 10 speakers. The proposed method is validated by performance estimation as well as classification accuracies in contradiction to other models.The proposed method gives better recognition rate and 93.33% accuracy is attained.     

基于顺序统计滤波的实时语音端点检测算法

针对嵌入式语音识别系统,提出了一种高效的实时语音端点检测算法. 算法以子带频谱熵为语音/噪声的区分特征, 首先将每帧语音的频谱划分成若干个子带, 计算出每个子带的频谱熵, 然后把相继若干帧的子带频谱熵经过一组顺序统计滤波器获得每帧的频谱熵, 根据频谱熵的值对输入的语音进行分类. 实验结果表明, 该算法能够有效地区分语音和噪声, 可以显著地提高语音识别系统的性能. 在不同的噪声环境和信噪比条件下具有鲁棒性. 此外, 本文提出的算法计算代价小, 简单易实现, 适合实时嵌入式语音识别系统的应用.     

Combination of features through weighted ensembles for image classification

Image classification is a multi-class problem that is usually tackled with ensembles of binary classifiers. Furthermore, one of the most important challenges in this field is to find a set of highly discriminative image features for reaching a good performance in image classification. In this work we propose to use weighted ensembles as a method for feature combination. First, a set of binary classifiers are trained with a set of features and then, the scores are weighted with distances obtained from another set of feature vectors. We present two different approaches to weight the score vector: (1) directly multiplying each score by the weights and (2) fusing the scores values and the distances through a Neural Network. The experiments have shown that the proposed methodology improves classification accuracy of simple ensembles and even more it obtains similar classification accuracy than state-of-the-art methods, but using much less parameters.     

A heterogeneous speech feature vectors generation approach with hybrid hmm classifiers

Automatic speech recognition (ASR) system plays a vital role in the human–machine interaction. ASR system faces the challenge of performance degradation due to inconsistency between training and testing phases. This occurs due to extraction and representation of erroneous, redundant feature vectors. This paper proposes three different combinations at speech feature vector generation phase and two hybrid classifiers at modeling phase. In feature extraction phase MFCC, RASTA-PLP, and PLP are combined in different ways. In modeling phase, the mean and variance are calculated to generate the inter and intra class feature vectors. These feature vectors are further adopted by optimization algorithm to generate refined feature vectors with traditional statistical technique. This approach uses GA?+?HMM and DE?+?HMM techniques to produce refine model parameters. The experiments are conducted on datasets of large vocabulary isolated Punjabi lexicons. The simulation result shows the performance improvement using MFCC and DE?+?HMM technique when compared with RASTA-PLP, PLP using hybrid HMM classifiers.     

Acoustic modeling problem for automatic speech recognition system: advances and refinements (Part II)

In automatic speech recognition (ASR) systems, hidden Markov models (HMMs) have been widely used for modeling the temporal speech signal. As discussed in Part I, the conventional acoustic models used for ASR have many drawbacks like weak duration modeling and poor discrimination. This paper (Part II) presents a review on the techniques which have been proposed in literature for the refinements of standard HMM methods to cope with their limitations. Current advancements related to this topic are also outlined. The approaches emphasized in this part of review are connectionist approach, explicit duration modeling, discriminative training and margin based estimation methods. Further, various challenges and performance issues such as environmental variability, tied mixture modeling, and handling of distant speech signals are analyzed along with the directions for future research.     

A hybrid multiple feature construction approach for classification using Genetic Programming

The purpose of feature construction is to create new higher-level features from original ones. Genetic Programming (GP) was usually employed to perform feature construction tasks due to its flexible representation. Filter-based approach and wrapper-based approach are two commonly used feature construction approaches according to their different evaluation functions. In this paper, we propose a hybrid feature construction approach using genetic programming (Hybrid-GPFC) that combines filter’s fitness function and wrapper’s fitness function, and propose a multiple feature construction method that stores top excellent individuals during a single GP run. Experiments on ten datasets show that our proposed multiple feature construction method (Fcm) can achieve better (or equivalent) classification performance than the single feature construction method (Fcs), and our Hybrid-GPFC can obtain better classification performance than filter-based feature construction approaches (Filter-GPFC) and wrapper-based feature construction approaches (Wrapper-GPFC) in most cases. Further investigations on combinations of constructed features and original features show that constructed features augmented with original features do not improve the classification performance comparing with constructed features only. The comparisons with three state-of-art methods show that in majority of cases, our proposed hybrid multiple feature construction approach can achieve better classification performance.     

Real-time lip reading system for isolated Korean word recognition

This paper proposes a real-time lip reading system (consisting of a lip detector, lip tracker, lip activation detector, and word classifier), which can recognize isolated Korean words. Lip detection is performed in several stages: face detection, eye detection, mouth detection, mouth end-point detection, and active appearance model (AAM) fitting. Lip tracking is then undertaken via a novel two-stage lip tracking method, where the model-based Lucas-Kanade feature tracker is used to track the outer lip, and then a fast block matching algorithm is used to track the inner lip. Lip activation detection is undertaken through a neural network classifier, the input for which being a combination of the lip motion energy function and the first dominant shape feature. In the last step, input words are defined and recognized by three different classifiers: HMM, ANN, and K-NN. We combine the proposed lip reading system with an audio-only automatic speech recognition (ASR) system to improve the word recognition performance in the noisy environments. We then demonstrate the potential applicability of the combined system for use within hands free in-vehicle navigation devices. Results from experiments undertaken on 30 isolated Korean words using the K-NN classifier at a speed of 15 fps demonstrate that the proposed lip reading system achieves a 92.67% word correct rate (WCR) for person-dependent tests, and a 46.50% WCR for person-independent tests. Also, the combined audio-visual ASR system increases the WCR from 0% to 60% in a noisy environment.     

Robust speech recognition using spatial-temporal feature distribution characteristics

Histogram equalization (HEQ) is one of the most efficient and effective techniques that have been used to reduce the mismatch between training and test acoustic conditions. However, most of the current HEQ methods are merely performed in a dimension-wise manner and without allowing for the contextual relationships between consecutive speech frames. In this paper, we present several novel HEQ approaches that exploit spatial-temporal feature distribution characteristics for speech feature normalization. The automatic speech recognition (ASR) experiments were carried out on the Aurora-2 standard noise-robust ASR task. The performance of the presented approaches was thoroughly tested and verified by comparisons with the other popular HEQ methods. The experimental results show that for clean-condition training, our approaches yield a significant word error rate reduction over the baseline system, and also give competitive performance relative to the other HEQ methods compared in this paper.     

一种基于自适应证据推理规则的集成学习方法

当前集成学习中的结合策略难以兼顾各个基学习器之间的信息和模型的可解释性。使用证据推理（evidential reasoning,ER）规则作为结合策略,将各个基学习器结果作为证据参与融合,可以较好地解决以上问题。但传统ER规则的证据参数是单一的,对不同的基学习器模型使用相同的证据参数显然是不合理的。为此,提出一种基于自适应证据推理（adaptive-evidential reasoning,A-ER）规则的集成学习方法,该方法在每次证据融合前对证据的类别进行判断,针对不同的证据类别自适应分配不同的证据参数。通过不同的分类案例表明,该方法与案例中其他方法相比具有更高的分类精度,证明了该方法使证据参数设置更加合理且具有更好的可解释性和泛化能力。     

脱机手写体汉字识别综述

脱机手写体汉字识别是模式识别领域中的难题之一。该文分析影响脱机手写体汉字识别性能的主要方面,如规范化方法、特征提取方法及分类方法,给出了每种方法的适用条件,介绍了目前研究中所使用的数据库。     

A case-based reasoning system with the two-dimensional reduction technique for customer classification

Many studies have tried to optimize parameters of case-based reasoning (CBR) systems. Among them, selection of appropriate features to measure similarity between the input and stored cases more precisely, and selection of appropriate instances to eliminate noises which distort prediction have been popular. However, these approaches have been applied independently although their simultaneous optimization may improve the prediction performance synergetically. This study proposes a case-based reasoning system with the two-dimensional reduction technique. In this study, vertical and horizontal dimensions of the research data are reduced through our research model, the hybrid feature and instance selection process using genetic algorithms. We apply the proposed model to a case involving real-world customer classification which predicts customers’ buying behavior for a specific product using their demographic characteristics. Experimental results show that the proposed technique may improve the classification accuracy and outperform various optimized models of the typical CBR system.     

An optimum feature extraction method based on Wavelet–Radon Transform and Dynamic Neural Network for pavement distress classification

Quantification of pavement crack data is one of the most important criteria in determining optimum pavement maintenance strategies. Recently, multi-resolution analysis such as wavelet decompositions provides very good multi-resolution analytical tools for different scales of pavement analysis and distresses classification. This paper present an automatic diagnosis system for detecting and classification pavement crack distress based on Wavelet–Radon Transform (WR) and Dynamic Neural Network (DNN) threshold selection. The algorithm of the proposed system consists of a combination of feature extraction using WR and classification using the neural network technique. The proposed WR + DNN system performance is compared with static neural network (SNN). In test stage; proposed method was applied to the pavement images database to evaluate the system performance. The correct classification rate (CCR) of proposed system is over 99%. This research demonstrated that the WR + DNN method can be used efficiently for fast automatic pavement distress detection and classification. The details of the image processing technique and the characteristic of system are also described in this paper.     

Environmental Independent ASR Model Adaptation/Compensation by Bayesian Parametric Representation

The mismatch between system training and operating conditions can seriously deteriorate the performance of automatic speech recognition (ASR) systems. Various techniques have been proposed to solve this problem in a specified speech environment. Employment of these techniques often involves modification on the ASR system structure. In this paper, we propose an environment-independent (EI) ASR model parameter adaptation approach based on Bayesian parametric representation (BPR), which is able to adapt ASR models to new environments without changing the structure of an ASR system. The parameter set of BPR is optimized by a maximum joint likelihood criterion which is consistent with that of the hidden Markov model (HMM)-based ASR model through an independent expectation-maximization (EM) procedure. Variations of the proposed approach are investigated in the experiments designed in two different speech environments: one is the noisy environment provided by the AURORA 2 database, and the other is the network environment provided by the NTIMIT database. Performances of the proposed EI ASR model compensation approach are compared to those of the cepstral mean normalization (CMN) approach, which is one of the standard techniques for additive noise compensation. The experimental results show that performances of ASR models in different speech environments are significantly improved after being adapted by the proposed BPR model compensation approach     

Phrase classes in two-level language models for ASR

In this work, we propose and compare two different approaches to a two-level language model. Both of them are based on phrase classes but they consider different ways of dealing with phrases into the classes. We provide a complete formulation consistent with the two approaches. The language models proposed were integrated into an Automatic Speech Recognition (ASR) system and evaluated in terms of Word Error Rate. Several series of experiments were carried out over a spontaneous human–machine dialogue corpus in Spanish, where users asked for information about long-distance trains by telephone. It can be extracted from the obtained results that the integration of phrases into classes when using the language models proposed leads to an improvement of the performance of an ASR system. Moreover, the obtained results seem to indicate that the history length with which the best performance is achieved is related to the features of the model itself. Thus, not all the models show the best results with the same value of history length.