On-line cursive script recognition using time-delay neural networks and hidden Markov models

In this paper, an evaluation of various discriminant neural networks classifiers for text- independent speaker verification problem is presented. Each person to be verified has a personalized neural network model. A new classifier called neural tree network (NTN) is also examined for this application. The memoryless feedforward neural network architecture makes decisions based on static features. Time delay neural network (TDNNs) have proved to be an efficient way to handle the dynamic nature of speech. Furthermore, a model called recurrent time delay neural networks (RTDNNs), obtained through a local feedback connection at the first hidden layer level of TDNNs is investigated. The training is carried out by backpropagation for sequence algorithm. The database used is a subset of the TIMIT database consisting of 38 speakers from the same dialect region. The NTN is compared with the MLP, TDNN, and RTDNN. It is shown that NTN is found to perform better than the other neural networks classifiers. Also, a little bit performance improvement was achieved due to the addition of temporal information for text-independent speaker verification problem using TDNNs and RTDNNs. Finally, we described the experimental results obtained using different neural network models.© (1994) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

This work presents survey, implementation and test for a neural network: TDNN (Time Delay Neural Network), applied to on-line handwritten recognition characters. In this work, we present a recognizer conception for on-line Arabic handwriting. On-line handwriting recognition of Arabic script is a complex problem, since it is naturally both cursive and unconstrained. This system permits to interpret a script represented by the pen trajectory. This technique is used notably in the electronic tablets. We will construct a data base with several scripters. Afterwards, and before attacking the recognition phase, there is a constructional samples phase of Arabic characters acquired from an electronic tablet to digitize (NOUN DATABASE). Obtained scores shows an effectiveness of the proposed approach based on convolutional neural networks.

This paper concerns with the online recognition of isolated hand writing Arabic characters, through, the interpretation of a script presented by a pen trajectory. This technique was generally used in the electronic organizers of Personal Digital Assistant type. First of all, we have built a data base with several scripters using a graphic tablet which will be used in our application. In order to have a precise recognition of the isolated characters, it is important to model their structure the most correctly possible. In this work we present the study, the implementation and the result of the test of a particular neural network which is the Time Delay Neural Networks. We have followed a two steps approach, in the first one, the character characteristics are extracted, and in the second one, a temporal multi-layered perception is developed for a future classification. Our temporal approach with the adaptive topology, responds to the nature of the Arabic script during the acquisition phase while the use of different learning algorithms can minimize the cost function and improve recognition rates. The parameterization of these two parts will allow us to analyze the impact of the neural network topology on the results of character recognition rates.

This paper reports the development of a multi-genre Urdu Broadcast (BC) corpus and a Large Vocabulary Continuous Speech Recognition (LVCSR) system. BC speech corpus of 98 hours from 453 speakers is collected and annotated. For acoustic modeling, Time-delay Neural Network (TDNN) is developed with prior Gaussian Mixture Model-Hidden Markov Model (GMM-HMM) training and alignments. For the language model, 3-gram, 4-gram and Recurrent Neural Network (RNN) based models are developed on a text corpus of 188 million words. The developed models are tested on 4.3 hours of unseen BC multi-genre speech dataset and the best Word Error Rate (WER) 18.59% is achieved using RNN based Language Model (LM). Moreover, a detailed word error analysis is carried out to compare the errors made by humans and the Automatic Speech Recognition (ASR) System. The results showed a similar behavior of word misrecognitions by both humans and ASR.

The use of Deep Neural Networks (DNN) shows significantly higher accuracy compared to traditional methods like the Hidden Markov Model (HMM) combined with the Gaussian Mixture Model (GMM) for creating acoustic models.This research involved developing and evaluating a baseline GMM-HMM model alongside a hybrid model that merges Time-Delay Neural Networks (TDNN) with LSTM and GMM-HMM for acoustic modelling, utilising the opensource Kaldi ASR toolkit.The main goal is to detect pronunciation errors of Arabic as spoken by Indians, and of the recitation of the Holy Qur'an, focusing specifically on ten Arabic letters () that non-Arabic speakers often mispronounce, confusing them with other letters that have similar articulation points.The speech dataset consisted of around 65 hours of audio, with 58 hours designated for training and 7 hours for validation and testing.The results indicate that the hybrid model, which combines TDNN-LSTM with GMM-HMM, achieved the highest performance of 96.88%, with a Word Error Rate (WER) of 3.12%.This outperforms the GMM-HMM model, which had a performance of 95.2% and a WER of 4.68%.These results confirm the hybrid model's effectiveness in improving the accuracy of identifying pronunciation errors in Indian speech and recitation compared to the GMM-HMM model alone.This represents a significant step forward in the development of more accurate and efficient speech recognition systems.

Recently, x-vector [1] has been a successful and popular approach for speaker verification, which employs a time delay neural network (TDNN) and statistics pooling to extract speaker characterizing embedding from variable-length utterances. Improvement upon the x-vector has been an active research area, and enormous neural networks have been elaborately designed based on the x-vector, e.g., extended TDNN (E-TDNN) [2], factorized TDNN (F-TDNN) [3], and densely connected TDNN (D-TDNN) [4]. In this work, we try to identify the optimal architectures from a TDNN based search space employing neural architecture search (NAS), named SpeechNAS. Leveraging the recent advances in the speaker recognition, such as high-order statistics pooling, multi-branch mechanism, D-TDNN and angular additive margin softmax (AAM) loss with a minimum hyper-spherical energy (MHE), SpeechNAS automatically discovers five network architectures, from SpeechNAS-1 to SpeechNAS-5, of various numbers of parameters and GFLOPs on the large-scale text-independent speaker recognition dataset VoxCelebl. Our derived best neural network achieves an equal error rate (EER) of 1.02% on the standard test set of VoxCelebl, which surpasses previous TDNN based state-of-the-art approaches by a large margin.

Neural networks with time delays offer the potential of providing massive parallelism and adaptation, and can provide excellent discrimination for temporal signal processing tasks. In a time delay neural network (TDNN), the shift-invariance capability required in speech recognition is achieved by explicitly making time-shifted copies of the inputs and internal responses and linking their corresponding weights. In this paper, we propose a limited feedback TDNN (LFB-TDNN), which feedbacks only those errors with significant magnitudes over a limited time span during the semi-batch back-propagation learning. Simulations show that the LFB-TDNN achieves favorable performance in isolated speech recognition tasks when compared with the standard TDNN, the discrete density and the continuous density hidden Markov models.

Sea state forecasting is of primary importance for marine operations related to offshore oil&gas industry. Operative decisions are based on marine weather forecasts. For this reason, at least once a day a bulletin of marine weather forecasts is delivered on board. Bulletins give a very good general overview of weather for the next days, while sometime they are not accurate relatively to the forecasted seastate values or their time of occurrence. In this work has been developed an Artificial Intelligence model named time delayed neural networks (TDNN) to improve the accuracy of short term sea state forecasting, combining bulletins forecasts with buoy measurements. The TDNN parameters are optimized by Genetic Algorithms. It has been found that the optimized TDNN produces very good forecasts, always outperforming bulletins, at least up to 24 hours ahead. Two configurations of TDNN are analyzed: in the first one, TDNN is trained with only buoy measurements, while in the second one TDNN is trained with both buoy and bulletin data. For 24 hours lead time, the combined use of buoy measurements and bulletins as input has been demonstrated to give better results than the use of only buoy measurements.

A method for ankle torque prediction ahead of the current time is proposed in this paper. The mean average value of EMG signals from four muscles, alongside the joint angle and angular velocity of the right ankle, were used as input parameters to train a time-delayed artificial neural network. Data collected from five healthy subjects were used to generate the dataset to train and test the model. The model predicted ankle torque for five different future times from zero to 2 seconds. Model predictions were compared to torque calculated from inverse dynamics for each subject. The model predicted ankle torque up to 1 second ahead of time with normalized root mean squared error of less than 15 percent while the coefficient of determination was over 0.85.Clinical Relevance- the potential of the model for predicting joint torque ahead of time is helpful to establish an intuitive interaction between human and assistive robots. This model has application to assist patients with neurological disorders.

This paper investigated the speaker independent vowel recognition for Malay children using the Time Delay Neural Network (TDNN). Due to less research done on the children speech recognition, the temporal structure of the children speech was not fully understood. This 2 hidden layers TDNN was proposed to discriminate 6 Malay vowels: /a/, /e/, /ə/, /i/, /o/ and /u/. The speech database consisted of vowel sounds from 360 children speakers. Cepstral coefficient was normalized for the input of TDNN. The frame rate of the TDNN was tested with 10ms, 20ms, and 30ms. It was found out that the 30ms frame rate produced the highest vowel recognition accuracy with 81.92%. The TDNN also showed higher speech recognition rate compared to the previous studies that used Multilayer Perceptron.

Prediction of hourly solar radiation using a novel hybrid model of ARMA and TDNN

Presented in this work is a possible myoelectric control scheme for a rehabilitation robotic application. The control input is from a time delayed neural network (TDNN). The input to the TDNN is four electromyographic (EMG) signals associated with the movement of the elbow and shoulder joints. The output of the TDNN is the joint position of the elbow and the joint position of the shoulder in the sagittal plane. The results presented here show the possibility of controlling multiple degrees of freedom at once. Prior work has shown that the optimal delay for accurate position prediction from a TDNN was 875ms with a 125ms interval, but this work shows that a delay of 300ms and a 100ms interval achieves similar results. This points to the feasibility of a TDNN based control scheme.

Cardiovascular pathologies cause 23.5% of human deaths, worldwide. An auto-diagnostic system monitoring heart activity, which can identify the early symptoms of cardiac illnesses, might reduce the death rate caused by these problems. Phonocardiography (PCG) is one of the possible techniques able to detect heart problems. Nevertheless, acoustic signal enhancement is required since it is exposed to various disturbances coming from different sources. The most common denoising enhancement is based on the Wavelet Transform (WT). However, the WT is highly susceptible to variations in the noise frequency distribution. This paper proposes a new adaptive denoising algorithm, which combines WT and Time Delay Neural Networks (TDNN). The acquired signal is decomposed by means of the WT using the coif five-wavelet basis at the tenth decomposition level and then provided as input to the TDNN. Besides the advantage of adaptive thresholding, the reason for using TDNNs is their capacity of estimating the Inverse Wavelet Transform (IWT). The best parameters of the TDNN were found for a NN consisting of 25 neurons in the first and 15 in the second layer and the delay block of 12 samples. The method was evaluated on several pathological heart sounds and on signals recorded in a noisy environment. The performance of the developed system with respect to other wavelet-based denoising approaches was validated by the online questionnaire.

The work is devoted to resolving the contradiction between the accuracy of modelling nonlinear dynamics and the speed of model construction under conditions of limited computing resources. The purposeof the work is to reduce the time for building time delay neural networks while ensuring a given accuracy in the tasks of identifying nonlinear dynamic objects with continuous characteristics. This goal is achieved by developing a method for pre-training neural networks that reflect the basic characteristics of the subject area. The scientific novelty of the work is the development of a method for identifying nonlinear dynamic objectsin the form of time delay neural networks based on the use of a set of basic pre-trained neural networks that reflect the typical properties of the subject area. In contrast to the traditional approach to pre-training, the developed method allows building models of lower complexity.A formal criterion is proposed for determining the moment of termination of the neural network pre-training, the use of which allows avoiding retraining of the base model and ensuring a significant reduction in the model training time on the target data set. The practical utility of the work lies in the development of an algorithm for the method of pre-training time delay neural networks in the tasks of identifying nonlinear dynamic objects with continuous characteristics, which allows to significantlyreduce the training time of neural networks without losing model accuracy. The value of this study is todetermine the area of effective use of the proposed method, namely, when the general and target datasets do not have significant differences and the target dataset is of sufficient size to reflect the properties of the research object.

Automatic speech recognition (ASR) technology has reached a mature level, and improving performance in data-scarce scenarios has become a key research focus. In this study, we propose a novel approach to building acoustic models by leveraging the multi-head attention mechanism to enhance various acoustic features extracted from audio data, combined with time-delay neural networks (TDNNs) using semi-orthogonal low-rank matrix factorization, resulting in significant performance improvements. To increase the robustness of the acoustic model, we introduce a small amount of data perturbation during the early stages of training. Additionally, during the decoding phase, we employ an external recurrent neural network language model (RNNLM) for rescoring, further enhancing the model’s accuracy. A comprehensive evaluation of the widely used LibriSpeech corpus shows that our method, using only half the training data, achieves word error rates (WERs) of 3.15 and 7.04 on the test_clean and test_other datasets, respectively. These results outperform models with similar architectures trained on the full dataset and demonstrate performance comparable to mainstream end-to-end models.