In this article, we propose a new method for joint cochannel speaker separation and recognition called adaptive-dictionary non-negative matrix deconvolution (DANMD). This method is an extension of non-negative matrix deconvolution (NMD) which models spectrogram matrix as a linear combination of dictionary elements (atoms). We propose a dictionary which is a linear combination of speaker-independent

In this paper, we propose a novel method to segment and label pass-phrase utterances for training deep neural network (DNN) bottleneck (BN) features for text-dependent speaker verification (TD-SV). Specifically, gender-dependent hidden Markov models (HMMs) for monophones are first trained using the pass-phrase utterances that are disjoint from evaluation. Next, the trained HMMs are speaker-adapted

The present work proposes a text-independent dialect identification system. Generally, dialects of a language exhibit varying pronunciation styles followed in a specific geographical region. In this paper, chroma features familiar with music-related systems are employed for identification of dialects. In addition, eight significant spectral shape related features from short term spectra are computed

Capsule networks (CapsNets) have recently gotten attention as a novel neural architecture. This paper presents the sequential routing framework which we believe is the first method to adapt a CapsNet-only structure to sequence-to-sequence recognition. Input sequences are capsulized then sliced by a window size. Each slice is classified to a label at the corresponding time through iterative routing

Glottal source characteristics vary between phonation types due to the tension of laryngeal muscles with the respiratory effort. Previous studies in the classification of phonation type have mainly used speech signals recorded by microphone. Recently, two studies were published in the classification of phonation type using neck surface accelerometer (NSA) signals. However, there are no previous studies

The vowel transition regions are the crucial landmarks in the speech signal. These vital regions are present at both ends of the vowel. They lie in the junction between a consonant and a vowel (CV) regions. This region plays an important role in numerous speech applications like speaker recognition, emotion conversion, speech rate modification, and CV unit recognition. The performance of these applications

Aspect-based sentiment analysis is a special type of sentiment analysis that aims to identify the discussed aspects and their sentiment polarities in a given review. In this paper, two deep learning models are proposed to address essential aspect-based sentiment analysis tasks: aspect-category identification and aspect-sentiment classification. For the first task, an identification model is proposed

In voice production the random variation of the glottal cycles in relation to a mean value caused by the (quasi-) periodic movement of the vocal folds generates the jitter phenomenon. Its study is important due to applications as identification of pathologies related to the voice, improvement of the naturalness of synthesized voices and calibration of signal processing algorithms to identify glottal

Speech encodes a wealth of information related to human behavior and has been used in a variety of automated behavior recognition tasks. However, extracting behavioral information from speech remains challenging including due to inadequate training data resources stemming from the often low occurrence frequencies of specific behavioral patterns. Moreover, supervised behavioral modeling typically relies

We investigate the predictive capability of mathematical models of the type-token relationship applied to the vocabulary growth profiles of selected of English language documents. We compare the existing Good-Toulmin and Heaps formulae with an alternative approach based on Bernoulli trial word selection from a fixed finite vocabulary using the Zipf and Zipf-Mandelbrot probability distributions. We

Anomia (word-finding difficulties) is the hallmark of aphasia, an acquired language disorder most commonly caused by stroke. Assessment of speech performance using picture naming tasks is a key method for both diagnosis and monitoring of responses to treatment interventions by people with aphasia (PWA). Currently, this assessment is conducted manually by speech and language therapists (SLT). Surprisingly

Nasalization of vowels is a phenomenon where the oral and nasal tracts participate simultaneously in the production of speech. Acoustic coupling of the oral and nasal tracts, together with the glottal activity, results in a complex dynamic production system. The extent of coupling determines the degree of nasalization in vowels. Changes in the instantaneous response of the vocal tract system are used

Mild Cognitive Impairment (MCI) is a heterogeneous clinical syndrome, often considered as the prodromal stage of dementia. It is characterized by the subtle deterioration of cognitive functions, including memory, executive functions and language. Mainly due to the tenuous nature of these impairments, a high percentage of MCI cases remain undetected. There is evidence that language changes in MCI are

Dysarthria is a condition which hampers the ability of an individual to control the muscles that play a major role in speech delivery. The loss of fine control over muscles that assist the movement of lips, vocal chords, tongue and diaphragm results in abnormal speech delivery. One can assess the severity level of dysarthria by analyzing the intelligibility of speech spoken by an individual. Continuous

While human listeners can, to some extent, understand the information conveyed by the speech signal when it is mixed with noise, traditional objective intelligibility measures usually fail to operate without a priori knowledge of the clean speech signal. This hence limits the usability of those measures in situations where the clean speech signal is inaccessible. In this paper a glimpse-based method

Recently, a variety of model designs and methods have blossomed in the context of the sentiment analysis domain. However, there is still a lack of comprehensive studies of Aspect-based Sentiment Analysis. We want to fill this gap and propose a comparison with ablation analysis of Aspect Term Extraction using various text embeddings methods. We particularly focused on simple architectures based on long

Heart failure (HF) is a major global health concern and is increasing in prevalence. It affects the larynx and breathing – thereby the quality of speech. In this article, we propose an approach for the automatic detection of people with HF using the speech signal. The proposed method explores mel-frequency cepstral coefficient (MFCC) features, glottal features, and their combination to distinguish

The pharyngeal fricative is a typical compensatory articulation disorder in cleft palate speech. It is produced by retracting the root of the tongue to the posterior pharyngeal wall to substitute for the fricatives and affricates produced in the oral cavity. People who use the pharyngeal fricative have difficulties in daily communication. Research on automatic pharyngeal fricative detection can provide

Representation transfer learning has been widely used across a range of machine learning tasks. One such notable approach seen in the speech literature is the use of Convolutional Neural Networks, pre-trained for image classification tasks, to extract features from spectrograms of speech signals. Interestingly, despite the strong performance of such approaches, there have been minimal research efforts

Far-field speech processing is an important and challenging problem. In this paper, we propose deep ad-hoc beamforming, a deep-learning-based multichannel speech enhancement framework based on ad-hoc microphone arrays, to address the problem. It contains three novel components. First, it combines ad-hoc microphone arrays with deep-learning-based multichannel speech enhancement, which reduces the probability

Robust speaker recognition, including in the presence of malicious attacks, is becoming increasingly important and essential, especially due to the proliferation of smart speakers and personal agents that interact with an individual’s voice commands to perform diverse and even sensitive tasks. Adversarial attack is a recently revived domain which is shown to be effective in breaking deep neural network-based

Prosody is the supra-segmental aspect of speech that helps to convey the structure and intended meaning of lexical content unambiguously. The automatic detection of prosodic events, such as phrase boundary and word prominence, has a number of applications in discourse analysis, where a combination of syntactic and acoustic-prosodic features is typically employed. This work addresses prosodic event

This paper presents a joint syntactic-semantic embedding model which not only uses syntactic information to enrich the word embeddings but also generates distributed representations for the syntactic structures themselves. The syntactic input to our model comes from a Lexicalized Tree-Adjoining Grammar parser. The word embeddings from our model outperform the Skip-gram embeddings in several word similarity

Alzheimer’s disease (AD) is one of the most common forms of dementia in the world. The Mini-Mental State Examination, a tool developed to detect AD, is composed of various tests that evaluate functional performance in several fields, one of which is language. Several symptoms are manifested in voices as a result of language and speech problems caused by AD, including frequent involuntary pauses during

Global vectors, or global embeddings, are important word representations for many natural language processing tasks. With the popularity of dynamic embeddings (also known as contextual embeddings, such as ELMo and BERT) in recent years, attentions on global vectors have been diverted to a large extent. While, compared to the dynamic embeddings, the global embeddings are faster to train, straightforward

Short text is the main carrier for people to express their ideas and opinions. And it is very important as well a big challenge to understand the meaning of short text or recognize the semantic patterns of different short texts. Most of existing methods use word embedding and short text interaction to learn short text pairs semantic patterns. However, some of these methods are complicated and cannot

This paper presents a textual entailment (TE) model that considers entailment as an optimization problem. The proposed TE model employs a genetic algorithm to derive an optimal similarity function and correlated entailment judgment threshold. The similarity function is formulated through a linear combination of text similarity measures and weights. Two text similarity measures are considered: cosine

Verbal fluency – i.e. the ability to name as many words of a given category as possible in a defined time interval – is an integral part of neuropsychological test batteries for the diagnosis of dementia. Verbal fluency can be easily administered and thus may also be implemented in computerized dementia screening tests. In the present study we sought to investigate the capability of phonemic verbal

Neural networks have been widely used in the field of text classification, and have achieved good results on various Chinese datasets. However, for long text classification, there are a lot of redundant information in text data, and some of the redundant information may involve other topic information, which makes long text classification challenging. To solve the above problems, this paper proposes

Neural Referring Expression Generation (REG) models have shown promising results in generating expressions which uniquely describe visual objects. However, current REG models still lack the ability to produce diverse and unambiguous referring expressions (REs). To address the lack of diversity, we propose generating a set of diverse REs, rather than one-shot REs. To reduce the ambiguity of referring

The accelerating flow of information we encounter around the world today makes many companies deploy speech recognition systems that, to an ever-growing extent, process data on-line rather than off-line. These systems, e.g., for real-time 24/7 broadcast transcription, often work with input-stream data containing utterances in more than one language. This multilingual data can correctly be transcribed

Early prediction of neurodegenerative disorders such as Alzheimer’s disease (AD) and related dementias may facilitate earlier access to medical and social supports. Further, detection of individuals with preclinical disease may help to enrich clinical trial populations for studies examining disease-modifying interventions. Changes in speech and language patterns may occur in the early stages of neurodegenerative

Intuitively, image- or video-based recommendations seem to be more reliable than those containing plain text, and these types of recommendations have recently become widely encouraged and commonly seen across opinion sharing platforms. Considering their potential for manipulation, graphs (e.g., images and videos) are more vulnerable to spam than scripts. However, most state-of-the-art solutions for

Text generation with auxiliary attributes, such as topics or sentiments, has made remarkable progress. However, high-quality labeled data is difficult to obtain for the large-scale corpus. Therefore, this paper focuses on social emotion ranking aiming to identify social emotions with different intensities evoked by online documents, which could be potentially beneficial for further controlled text

Neural sequence-to-sequence text-to-speech synthesis (TTS) can produce high-quality speech directly from text or simple linguistic features such as phonemes. Unlike traditional pipeline TTS, the neural sequence-to-sequence TTS does not require manually annotated and complicated linguistic features such as part-of-speech tags and syntactic structures for system training. However, it must be carefully

The taking of turns is a fundamental aspect of dialogue. Since it is difficult to speak and listen at the same time, the participants need to coordinate who is currently speaking and when the next person can start to speak. Humans are very good at this coordination, and typically achieve fluent turn-taking with very small gaps and little overlap. Conversational systems (including voice assistants and

Headline generation is a task where the most important information of a news article is condensed and embodied into a single short sentence. This task is normally addressed by summarization techniques, ideally combining extractive and abstractive methods together with sentence compression or fusion techniques. Although Natural Language Generation (NLG) techniques have not been directly exploited for

Generating structured query language (SQL) from natural language is an emerging research topic. This paper presents a new learning paradigm from indirect supervision of the answers to natural language questions, instead of SQL queries. This paradigm facilitates the acquisition of training data due to the abundant resources of question-answer pairs for various domains in the Internet, and expels the

Currently, there is little agreement as to how Natural Language Generation (NLG) systems should be evaluated, with a particularly high degree of variation in the way that human evaluation is carried out. This paper provides an overview of how (mostly intrinsic) human evaluation is currently conducted and presents a set of best practices, grounded in the literature. These best practices are also linked

Visual storytelling is a creative and challenging task, aiming to automatically generate a story-like description for a sequence of images. The descriptions generated by previous visual storytelling approaches lack coherence because they use word-level sequence generation methods and do not adequately consider sentence-level dependencies. To tackle this problem, we propose a novel hierarchical visual

A laryngectomee is a person whose larynx has been removed by surgery, usually due to laryngeal cancer. After surgery, most laryngectomees are able to speak again, using techniques that are learned with the help of a speech therapist. This is termed as alaryngeal speech, and esophageal speech (ES) is one of the several alaryngeal speech production modes. A considerable amount of research has been dedicated

Existing methods for question answering over knowledge bases (KBQA) ignore the consideration of the model prediction uncertainties. We argue that estimating such uncertainties is crucial for the reliability and interpretability of KBQA systems. Therefore, we propose a novel end-to-end KBQA model based on Bayesian Neural Network (BNN) to estimate uncertainties arose from both model and data. To our

Query-based document summarization aims to extract or generate a summary of a document which directly answers or is relevant to the search query. It is an important technique that can be beneficial to a variety of applications such as search engines, document-level machine reading comprehension, and chatbots. Currently, datasets designed for query-based summarization are short in numbers and existing

Neural network methods which leverage word-embedding obtained from unsupervised learning models have been widely adopted in many natural language processing (NLP) tasks, including sentiment analysis and sentence classification. Existing sentence representation generation approaches which serve for classification tasks generally rely on complex deep neural networks but relatively simple loss functions

End-to-end spoken language translation (SLT) has recently gained popularity thanks to the advancement of sequence to sequence learning in its two parent tasks: automatic speech recognition (ASR) and machine translation (MT). However, research in the field has to confront with the scarcity of publicly available corpora to train data-hungry neural networks. Indeed, while traditional cascade solutions

The task of quantifying human behavior by observing interaction cues is an important and useful one across a range of domains in psychological research and practice. Machine learning-based approaches typically perform this task by first estimating behavior based on cues within an observation window, such as a fixed number of words, and then aggregating the behavior over all the windows in that interaction

This paper presents novel neural network based language models that can correct automatic speech recognition (ASR) errors by using speech recognizer outputs as a context. Our proposed models, called neural candidate-aware language models (NCALMs), estimate the generative probability of a target sentence while considering ASR outputs including hypotheses and their posterior probabilities. Recently,

Spoofing attacks have been acknowledged as a serious threat to automatic speaker verification (ASV) systems. In this paper, we are specifically concerned with replay attack scenarios. As a countermeasure to the problem, we propose a front-end based on the blind estimation of the channel response magnitude and as a back-end a residual neural network. Our hypothesis is that the magnitude response of

Sentence pair modelling is defined as the task of identifying the semantic interaction between a sentence pair, i.e., paraphrase and textual entailment identification and semantic similarity measurement. It constitutes a set of crucial tasks for research in the area of natural language understanding. Sentence representation learning is a fundamental technology for sentence pair modelling, where the

We propose a data-to-document generator that can easily control the contents of output texts based on a neural language model. Conventional data-to-text model is useful when a reader seeks a global summary of data because it has only to describe an important part that has been extracted beforehand. However, since it differs from users to users what they are interested in, it is necessary to develop

We describe a novel way to implement subword language models in speech recognition systems based on weighted finite state transducers, hidden Markov models, and deep neural networks. The acoustic models are built on graphemes in a way that no pronunciation dictionaries are needed, and they can be used together with any type of subword language model, including character models. The advantages of short

Replay attacks pose the most severe threat to automatic speaker verification systems among various spoofing attacks. In this paper, we propose a novel feature extraction method that leverages both the phase-based and magnitude-based features. The proposed method fully utilizes the subband information and the complementary information from the phase and magnitude spectra. First, we conduct a discriminative

Acoustic and articulatory differences between spoken and shouted vowels were analyzed for two male and two female subjects by means of acoustic recordings and midsagittal magnetic resonance images of the vocal tract. In accordance with previous acoustic findings, the fundamental frequencies, intensities, and formant frequencies were all generally higher for shouted than for spoken vowels. The harmonics-to-noise

The real-time Magnetic Resonance Imaging (rtMRI) video captures the vocal tract movements in the mid-sagittal plane during speech. Air tissue boundaries (ATBs) are contours that trace the transition between the high-intensity tissue corresponding to the speech articulators and the low-intensity airway cavity in the rtMRI video. The ATB segmentation in an rtMRI video is a common preprocessing step which

We present a novel abstractive document summarization based on the recently proposed dynamic convolutional encoder-decoder architectures. We address several aspects of summarization that are not well modeled by the basic architecture, by integrating multiple layers of the encoder, controlling information flow in the hierarchy, and exploiting external knowledge. First, we propose a simple and efficient

Web search users are likely to face problems related to the availability of large amounts of data. As the quantity of online content grows, the risk of missing relevant information during search can only increase. Moreover, external variables such as the users’ reading proficiency level can further complicate the task. This article proposes an online multi-document summarization algorithm for text

We study class-based n-gram and neural network language models for very large vocabulary speech recognition of two morphologically rich languages: Finnish and Estonian. Due to morphological processes such as derivation, inflection and compounding, the models need to be trained with vocabulary sizes of several millions of word types. Class-based language modelling is in this case a powerful approach

Designing a Spoken Language Understanding (SLU) system for command-and-control applications is challenging. Both Automatic Speech Recognition and Natural Language Understanding are language and application dependent to a great extent. Even with a lot of design effort, users often still have to know what to say to the system for it to do what they want. We propose to use an end-to-end SLU system that

The vulnerability of Automatic Speaker Verification (ASV) systems to spoofing or presentation attacks is still an open security issue. In this context, replay spoofing attacks pose a great threat to an ASV system since they can be easily performed (using a playback device, and without needing any technical skill). In this paper, we analyze replay speech signals in terms of reverberation that may occur