In recent years, deep learning-based speech synthesis has attracted a lot of attention from the machine learning and speech communities. In this paper, we propose Mixture-TTS, a non-autoregressive speech synthesis model based on mixture alignment mechanism. Mixture-TTS aims to optimize the alignment information between text sequences and mel-spectrogram. Mixture-TTS uses a linguistic encoder based on soft phoneme-level alignment and hard word-level alignment approaches, which explicitly extract word-level semantic information, and introduce pitch and energy predictors to optimally predict the rhythmic information of the audio. Specifically, Mixture-TTS introduces a post-net based on a five-layer 1D convolution network to optimize the reconfiguration capability of the mel-spectrogram. We connect the output of the decoder to the post-net through the residual network. The mel-spectrogram is converted into the final audio by the HiFi-GAN vocoder. We evaluate the performance of the Mixture-TTS on the AISHELL3 and LJSpeech datasets. Experimental results show that Mixture-TTS is somewhat better in alignment information between the text sequences and mel-spectrogram, and is able to achieve high-quality audio. The ablation studies demonstrate that the structure of Mixture-TTS is effective. Full article

The explosive volume of semantic data published in the Resource Description Framework (RDF) data model demands efficient management and compression with better compression ratio and runtime. Although extensive work has been carried out for compressing the RDF datasets, they do not perform well in all dimensions. However, these compressors rarely exploit the graph patterns and structural regularities of real-world datasets. Moreover, there are a variety of existing approaches that reduce the size of a graph by using a grammar-based graph compression algorithm. In this study, we introduce a novel approach named gRDF (graph repair for RDF) that uses gRePair, one of the most efficient grammar-based graph compression schemes, to compress the RDF dataset. In addition to that, we have improved the performance of HDT (header-dictionary-triple), an efficient approach for compressing the RDF datasets based on structural properties, by introducing modified HDT (M-HDT). It can detect the frequent graph pattern by employing the data-structure-oriented approach in a single pass from the dataset. In our proposed system, we use M-HDT for indexing the nodes and edge labels. Then, we employ gRePair algorithm for identifying the grammar from the RDF graph. Afterward, the system improves the performance of $k^2$-trees by introducing a more efficient algorithm to create the trees and serialize the RDF datasets. Our experiments affirm that the proposed gRDF scheme can substantially achieve at approximately 26.12%, 13.68%, 6.81%, 2.38%, and 12.76% better compression ratio when compared with the most prominent state-of-the-art schemes such as HDT, HDT++, $k^2$-trees, RDF-TR, and gRePair in the case of real-world datasets. Moreover, the processing efficiency of our proposed scheme also outperforms others. Full article

Successful applications of deep learning technologies in the natural language processing domain have improved text-based intent classifications. However, in practical spoken dialogue applications, the users’ articulation styles and background noises cause automatic speech recognition (ASR) errors, and these may lead language models to misclassify users’ intents. To overcome the limited performance of the intent classification task in the spoken dialogue system, we propose a novel approach that jointly uses both recognized text obtained by the ASR model and a given labeled text. In the evaluation phase, only the fine-tuned recognized language model (RLM) is used. The experimental results show that the proposed scheme is effective at classifying intents in the spoken dialogue system containing ASR errors. Full article

The expression of emotions in human communication plays a very important role in the information that needs to be conveyed to the partner. The forms of expression of human emotions are very rich. It could be body language, facial expressions, eye contact, laughter, and tone of voice. The languages of the world’s peoples are different, but even without understanding a language in communication, people can almost understand part of the message that the other partner wants to convey with emotional expressions as mentioned. Among the forms of human emotional expression, the expression of emotions through voice is perhaps the most studied. This article presents our research on speech emotion recognition using deep neural networks such as CNN, CRNN, and GRU. We used the Interactive Emotional Dyadic Motion Capture (IEMOCAP) corpus for the study with four emotions: anger, happiness, sadness, and neutrality. The feature parameters used for recognition include the Mel spectral coefficients and other parameters related to the spectrum and the intensity of the speech signal. The data augmentation was used by changing the voice and adding white noise. The results show that the GRU model gave the highest average recognition accuracy of 97.47%. This result is superior to existing studies on speech emotion recognition with the IEMOCAP corpus. Full article

This work presents a new approach based on a spiking neural network for sound preprocessing and classification. The proposed approach is biologically inspired by the biological neuron’s characteristic using spiking neurons, and Spike-Timing-Dependent Plasticity (STDP)-based learning rule. We propose a biologically plausible sound classification framework that uses a Spiking Neural Network (SNN) for detecting the embedded frequencies contained within an acoustic signal. This work also demonstrates an efficient hardware implementation of the SNN network based on the low-power Spike Continuous Time Neuron (SCTN). The proposed sound classification framework suggests direct Pulse Density Modulation (PDM) interfacing of the acoustic sensor with the SCTN-based network avoiding the usage of costly digital-to-analog conversions. This paper presents a new connectivity approach applied to Spiking Neuron (SN)-based neural networks. We suggest considering the SCTN neuron as a basic building block in the design of programmable analog electronics circuits. Usually, a neuron is used as a repeated modular element in any neural network structure, and the connectivity between the neurons located at different layers is well defined. Thus, generating a modular Neural Network structure composed of several layers with full or partial connectivity. The proposed approach suggests controlling the behavior of the spiking neurons, and applying smart connectivity to enable the design of simple analog circuits based on SNN. Unlike existing NN-based solutions for which the preprocessing phase is carried out using analog circuits and analog-to-digital conversion, we suggest integrating the preprocessing phase into the network. This approach allows referring to the basic SCTN as an analog module enabling the design of simple analog circuits based on SNN with unique inter-connections between the neurons. The efficiency of the proposed approach is demonstrated by implementing SCTN-based resonators for sound feature extraction and classification. The proposed SCTN-based sound classification approach demonstrates a classification accuracy of 98.73% using the Real-World Computing Partnership (RWCP) database. Full article

Given the excessive foul language identified in audio and video files and the detrimental consequences to an individual’s character and behaviour, content censorship is crucial to filter profanities from young viewers with higher exposure to uncensored content. Although manual detection and censorship were implemented, the methods proved tedious. Inevitably, misidentifications involving foul language owing to human weariness and the low performance in human visual systems concerning long screening time occurred. As such, this paper proposed an intelligent system for foul language censorship through a mechanized and strong detection method using advanced deep Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) through Long Short-Term Memory (LSTM) cells. Data on foul language were collected, annotated, augmented, and analysed for the development and evaluation of both CNN and RNN configurations. Hence, the results indicated the feasibility of the suggested systems by reporting a high volume of curse word identifications with only 2.53% to 5.92% of False Negative Rate (FNR). The proposed system outperformed state-of-the-art pre-trained neural networks on the novel foul language dataset and proved to reduce the computational cost with minimal trainable parameters. Full article

In this paper, we propose a multi-channel cross-tower with attention mechanisms in latent domain network (Multi-TALK) that suppresses both the acoustic echo and background noise. The proposed approach consists of the cross-tower network, a parallel encoder with an auxiliary encoder, and a decoder. For the multi-channel processing, a parallel encoder is used to extract latent features of each microphone, and the latent features including the spatial information are compressed by a 1D convolution operation. In addition, the latent features of the far-end are extracted by the auxiliary encoder, and they are effectively provided to the cross-tower network by using the attention mechanism. The cross tower network iteratively estimates the latent features of acoustic echo and background noise in each tower. To improve the performance at each iteration, the outputs of each tower are transmitted as the input for the next iteration of the neighboring tower. Before passing through the decoder, to estimate the near-end speech, attention mechanisms are further applied to remove the estimated acoustic echo and background noise from the compressed mixture to prevent speech distortion by over-suppression. Compared to the conventional algorithms, the proposed algorithm effectively suppresses the acoustic echo and background noise and significantly lowers the speech distortion. Full article

Speech emotion recognition (SER) plays a significant role in human–machine interaction. Emotion recognition from speech and its precise classification is a challenging task because a machine is unable to understand its context. For an accurate emotion classification, emotionally relevant features must be extracted from the speech data. Traditionally, handcrafted features were used for emotional classification from speech signals; however, they are not efficient enough to accurately depict the emotional states of the speaker. In this study, the benefits of a deep convolutional neural network (DCNN) for SER are explored. For this purpose, a pretrained network is used to extract features from state-of-the-art speech emotional datasets. Subsequently, a correlation-based feature selection technique is applied to the extracted features to select the most appropriate and discriminative features for SER. For the classification of emotions, we utilize support vector machines, random forests, the k-nearest neighbors algorithm, and neural network classifiers. Experiments are performed for speaker-dependent and speaker-independent SER using four publicly available datasets: the Berlin Dataset of Emotional Speech (Emo-DB), Surrey Audio Visual Expressed Emotion (SAVEE), Interactive Emotional Dyadic Motion Capture (IEMOCAP), and the Ryerson Audio Visual Dataset of Emotional Speech and Song (RAVDESS). Our proposed method achieves an accuracy of 95.10% for Emo-DB, 82.10% for SAVEE, 83.80% for IEMOCAP, and 81.30% for RAVDESS, for speaker-dependent SER experiments. Moreover, our method yields the best results for speaker-independent SER with existing handcrafted features-based SER approaches. Full article

The advent of new devices, technology, machine learning techniques, and the availability of free large speech corpora results in rapid and accurate speech recognition. In the last two decades, extensive research has been initiated by researchers and different organizations to experiment with new techniques and their applications in speech processing systems. There are several speech command based applications in the area of robotics, IoT, ubiquitous computing, and different human-computer interfaces. Various researchers have worked on enhancing the efficiency of speech command based systems and used the speech command dataset. However, none of them catered to noise in the same. Noise is one of the major challenges in any speech recognition system, as real-time noise is a very versatile and unavoidable factor that affects the performance of speech recognition systems, particularly those that have not learned the noise efficiently. We thoroughly analyse the latest trends in speech recognition and evaluate the speech command dataset on different machine learning based and deep learning based techniques. A novel technique is proposed for noise robustness by augmenting noise in training data. Our proposed technique is tested on clean and noisy data along with locally generated data and achieves much better results than existing state-of-the-art techniques, thus setting a new benchmark. Full article

Voice is one of the essential mechanisms for communicating and expressing one’s intentions as a human being. There are several causes of voice inability, including disease, accident, vocal abuse, medical surgery, ageing, and environmental pollution, and the risk of voice loss continues to increase. Novel approaches should have been developed for speech recognition and production because that would seriously undermine the quality of life and sometimes leads to isolation from society. In this review, we survey mouth interface technologies which are mouth-mounted devices for speech recognition, production, and volitional control, and the corresponding research to develop artificial mouth technologies based on various sensors, including electromyography (EMG), electroencephalography (EEG), electropalatography (EPG), electromagnetic articulography (EMA), permanent magnet articulography (PMA), gyros, images and 3-axial magnetic sensors, especially with deep learning techniques. We especially research various deep learning technologies related to voice recognition, including visual speech recognition, silent speech interface, and analyze its flow, and systematize them into a taxonomy. Finally, we discuss methods to solve the communication problems of people with disabilities in speaking and future research with respect to deep learning components. Full article