It is well known that the unpredictable speech production brought on by stress from the task at hand has a significant negative impact on the performance of speech processing algorithms. Speech therapy benefits from being able to detect stress in speech. Speech processing performance suffers noticeably when perceptually produced stress causes variations in speech production. Using the acoustic speech signal to objectively characterize speaker stress is one method for assessing production variances brought on (...) by stress. Real-world complexity and ambiguity make it difficult for decision-makers to express their conclusions with clarity in their speech. In particular, the Neutrosophic speech algorithm is used to encode the language variables because they cannot be computed directly. Neutrosophic sets are used to manage indeterminacy in a practical situation. Existing algorithms are used except for stress on Neutrosophic speech recognition. The creation of algorithms that calculate, categorize, or differentiate between different stress circumstances. Understanding stress and developing strategies to combat its effects on speech recognition and human-computer interaction system are the goals of this recognition. (shrink)

Speech recognition technology has played an indispensable role in realizing human-computer intelligent interaction. However, most of the current Chinese speech recognition systems are provided online or offline models with low accuracy and poor performance. To improve the performance of offline Chinese speech recognition, we propose a hybrid acoustic model of deep convolutional neural network, long short-term memory, and deep neural network. This model utilizes DCNN to reduce frequency variation and adds a batch normalization (...) layer after its convolutional layer to ensure the stability of data distribution, and then use LSTM to effectively solve the gradient vanishing problem. Finally, the fully connected structure of DNN is utilized to efficiently map the input features into a separable space, which is helpful for data classification. Therefore, leveraging the strengths of DCNN, LSTM, and DNN by combining them into a unified architecture can effectively improve speech recognition performance. Our model was tested on the open Chinese speech database THCHS-30 released by the Center for Speech and Language Technology of Tsinghua University, and it was concluded that the DLD model with 3 layers of LSTM and 3 layers of DNN had the best performance, reaching 13.49% of words error rate. (shrink)

With the development of modern information technology, the flipped classroom teaching mode came into being. It has gradually become one of the hotspots of contemporary educational circles and has been applied to various disciplines at the same time. The domestic research on the flipped classroom teaching mode is still in the exploratory stage. The application of flipped classroom teaching mode is still in the exploratory stage. It also has many problems, such as low class efficiency, poor teacher-student interaction, outdated (...) teaching modes, not student-centered, etc., which lead to poor students’ enthusiasm for learning. Therefore, the current English flipped classroom teaching mode still needs to be tested and revised in practice. Combined with emotion recognition technology, this paper analyzes speech emotion recognition, image emotion recognition, and audition emotion recognition technology and conducts a revision test for the current English flipped classroom teaching mode. It uses the SVM algorithm for one-to-one method and dimension discretization for emotion recognition, and finds that the recognition results after different dimension classification recognition are improved for each emotion. Among them, the recognition rate of different dimension classification recognition methods is 2.6% higher than that of one-to-one method. This shows that under the same conditions, the emotion recognition technology of different dimension classification recognition methods is higher. (shrink)

The first of its kind to thoroughly cover language technology at all levels and with all modern technologies this book takes an empirical approach to the ...

Locus equations contain an economical set of hidden (i.e., not directly observable in the data) parameters of speech that provide an elegant way of characterizing the ubiquitous context-dependent behaviors exhibited in speech acoustics. These hidden parameters can be effectively exploited to constrain the huge set of context-dependent speech model parameters currently in use in modern, mainstream speech recognition technology.

This paper outlines the background development of “intelligent“ technologies such as speech recognition. Despite significant progress in the development of these technologies, they still fall short in many areas, and rapid advances in areas such as dictation are actually stalled. In this paper we have proposed semi-automatic solutions — smart integration of human and intelligent efforts. One such technique involves improvement to the speech recognition editing interface, thereby reducing the perception of errors to the viewer. Other (...) techniques that are described in the paper are batch enrollment, which allows the user to reduce the amount of time required for enrollment, and content spotting, which can be used for applications that have repeated content flow, such as movies or museum tours. The paper also suggests a general concept of distributive training of speech recognition systems that is based on data collection across a network. (shrink)

To solve a series of pension problems caused by aging, based on the emotional recognition of the Internet of Things, the control method and system research of smart homes are proposed. This article makes a detailed analysis and research on the necessity, feasibility, and how to realize speech emotion recognition technology in smart families, introduces the definition and classification of emotion, and puts forward five main emotions to be recognized in speech emotion recognition based (...) on smart family environment. Then, based on this, it analyses the acquisition methods of emotional speech data. On this premise, this article discusses and analyses the related problems of voice data acquisition in smart homes, such as the voice characteristics and acquisition methods, puts forward three rules for voice text design, and determines the relatively suitable hybrid recording acquisition method applied in a smart home environment. At the same time, the design and establishment process of intelligent family emotional speech database is described in detail. The related problems of feature extraction in speech emotion recognition are studied. Starting from the definition of feature extraction, this article expounds on the necessity of feature extraction in the process of recognition and analyses the characteristics of the speech signals. For the specific environment of the smart family, the speech signal required to be processed needs to be close to the auditory characteristics of the human ears, and the speech signal contains enough information. Finally, the Mel frequency cepstrum coefficient is selected as the feature parameter applied in this article, and the extraction process of MFCC is introduced in detail. (shrink)

As development of the speech recognition system entirely depends upon the spoken language used for its development, and the very fact that speech technology is highly language dependent and reverse engineering is not possible, there is an utmost need to develop such systems for Indian languages. In this paper we present the implementation of a time delay neural network system (TDNN) in a modular fashion by exploiting the hidden structure of previously phonetic subcategory network for (...) class='Hi'>recognition of Hindi consonants. For the present study we have selected all the Hindi phonemes for srecognition. A vocabulary of 207 Hindi words was designed for the task-specific environment and used as a database. For the recognition of phoneme, a three-layered network was constructed and the network was trained using the back propagation learning algorithm. Experiments were conducted to categorize the Hindi voiced, unvoiced stops, semi vowels, vowels, nasals and fricatives. A close observation of confusion matrix of Hindi stops revealed maximum confusion of retroflex stops with their non-retroflex counterparts. (shrink)

What type of speech act is a norm of action, when the norm is agreed upon as the conclusion of an argumentative dialogue? My hypothesis is that, whenever a norm of action is the conclusion of an argument, it should be analyzed as the statement of a norm and thus as a verdictive speech act. If the context is appropriate, and the interlocutors are sincerely (or institutionally) committed to their argumentative exchange and its conclusion, then this verdictive motivates (...) and institutes a new one with the force of an exercitive. The interlocutors’ recognition and acceptance that the new illocution has been performed lends the norm its exercitive force. (shrink)

Oral English, as a language tool, is not only an important part of English learning but also an essential part. For nonnative English learners, effective and meaningful voice feedback is very important. At present, most of the traditional recognition and error correction systems for oral English training are still in the theoretical stage. At the same time, the corresponding high-end experimental prototype also has the disadvantages of large and complex system. In the speech recognition technology, the (...) traditional speech recognition technology is not perfect in recognition ability and recognition accuracy, and it relies too much on the recognition of speech content, which is easily affected by the noise environment. Based on this, this paper will develop and design a spoken English assistant pronunciation training system based on Android smartphone platform. Based on the in-depth study and analysis of spoken English speech correction algorithm and speech feedback mechanism, this paper proposes a lip motion judgment algorithm based on ultrasonic detection, which is used to assist the traditional speech recognition algorithm in double feedback judgment. In the feedback mechanism of intelligent speech training, a double benchmark scoring mechanism is introduced to comprehensively evaluate the speech of the speech trainer and correct the speaker’s speech in time. The experimental results show that the speech accuracy of the system reaches 85%, which improves the level of oral English trainers to a certain extent. (shrink)

The last decade has seen significant improvements in Artificial Intelligence (AI) technologies, including robotics, machine vision, speech recognition and text generation. Increasing automation will undoubtedly affect the future of work, and discussions on how the development of AI in the workplace will impact labor markets often include two scenarios: (1) labor replacement and (2) labor enabling. The former involves replacing workers with machines, while the latter assumes that human-machine cooperation can significantly improve worker productivity. In this context, it (...) is often argued that (1) could lead to mass unemployment and that (2) therefore would be more desirable. We argue, however, that the labor-enabling scenario conflates two distinct possibilities. On the one hand, technology can increase productivity while also promoting “the goods of work,” such as the opportunity to pursue excellence, experience a sense of community and contribute to society (human augmentation). On the other hand, higher productivity can also be achieved in a way that reduces opportunities for the “goods of work” and/or increases “the bads of work,” such as injury, reduced physical and mental health, reduction of autonomy, privacy and human dignity (human stunting). We outline the differences of these outcomes and discuss the implications for the labor market in the context of contemporaneous discussions on the value of work and human wellbeing. (shrink)

Speech recognition is an interdisciplinary subfield of natural language processing (NLP) that facilitates the recognition and translation of spoken language into text by machine. Speech recognition plays an important role in digital transformation. It is widely used in different areas such as education, industry, and healthcare and has recently been used in many Internet of Things and Machine Learning applications. The process of speech recognition is one of the most difficult processes in computer (...) science. Despite numerous searches in this domain, an optimal method for speech recognition has not yet been found. This is due to the fact that there are many attributes that characterize natural languages and every language has its particular highlights. The aim of this research is to provide a comprehensive understanding of the various techniques within the domain of Speech Recognition through a systematic literature review of the existing work. We will introduce the most significant and relevant techniques that may provide some directions in the future research. (shrink)

This article aims to present the problem of the impact of artificial intelligence on respect for human dignity in the sphere of care for people who, for various reasons, are described as particularly vulnerable, especially seniors and people with various disabilities. In recent years, various initiatives and works have been undertaken on the European scene to define the directions in which the development and use of artificial intelligence should go. According to the human-centric approach, artificial intelligence should be developed, used (...) and monitored with people in mind, their needs and rights. It is artificial intelligence that should adapt to the rules set by people, not the other way around. The source of all human rights and freedoms is respect for human dignity. This is evidenced by numerous European, international, national regulations and documents. Respecting this value is also one of the works on the AI development. One of the areas of our life into which AI enters more and more boldly and which in the near future, due to demographic reasons and aging population, may be almost dominated by modern technologies, is the medical and care system. The usefulness of industrial co-robots made us think about the possibility of using similar devices in the non-industrial sphere, including in the sphere of care, nursing and therapy. Actually, despite the impressive development of technology and AI, intelligent robots/devices used for care, nursing or therapy (so-called care/nursing or therapy robots) perform rather simple activities. They provide information, navigate, bring something and give it (e.g. medicines, food), help to get up. They are an important support for medical staff in this respect. Performing tasks that are more complex in terms of technology and movement, such as feeding, washing, intimate hygiene, dressing or undressing, is still beyond the reach of the currently used robots. The question we must ask ourselves is whether, in the future, care robots will be able to perform such tasks, and should they do so? Will this lead to the replacement of human medical personnel in the performance of the above-mentioned activities with such robots. Would anyone want to be "looked after" by a soulless machine? Should potential wards/patients, in other words all of us in fact, have the right to object to being taken into the care of an intelligent robot? There is no doubt that intelligent robots have great potential, especially when it comes to ensuring or maintaining the independence or mobility of people with various health limitations. However, we must not forget about the risks and dangers they may generate. Even a simple robot generates serious legal and ethical problems. Who should be liable for damage caused in connection with improper care performed with the participation of a robot? Robots based on AI systems are able to collect a significant amount of sensitive data, for example through the face and speech recognition function. There is considerable scope for abuse in this area, not only in terms of personal data protection but also in terms of informational self-determination, privacy and dignity. Bearing in mind the above risks, the creation of a new right to object to being looked after by a robot should be considered. (shrink)

Cochlear implants (CI) support the development of oral language in hearing-impaired children. However, even with CI, speech recognition in noise (SRiN) is limited. This raised the question, whether these restrictions are related to the quality of life (QoL) of children and adolescents with CI and how SRiN and QoL are related to each other. As a result of a systematic literature research only three studies were found, indicating positive moderating effects between SRiN and QoL of young CI users. (...) Thirty studies addressed the quality of life of children and adolescents with CI. Following the criteria of the World Health Organization (WHO) for pediatric HRQoL (1994) only a minority used validated child centered and age appropriate QoL instruments. Moreover, despite the consensus that usually children and adolescents are the most prominent informants of their own QoL (parent- reports complement the information of the children) only a minority of investigators used self-reports. Restricted SRiN may be a burden for the QoL of children and adolescents with CI. Up to now the CI community does not seem to be aware of a possible impairment of QoL in young CI users. Further studies addressing this topic are urgently needed, which is also relevant for parents, clinicians, therapists, teachers, and politicians. Additionally investigators should use valid pediatric QoL instruments. They should be aware, that most of the young CI users are able to inform about their quality of life also themselves (additionally to parents). (shrink)

Integrated design of financial self-service terminal based on artificial intelligence voice interaction with the rapid development of science and technology, artificial intelligence technology is deepening in the field of intelligence and automation. The financial industry is the lifeblood of a country’s economy, with great growth potential and high growth rate. The integrated design of intelligent financial self-service terminal has become an important topic in the field of rapid development of social economy and science and technology. Therefore, this (...) paper designs the integration of financial self-service terminal based on artificial intelligence voice interaction. First, this paper introduces the meaning and composition of financial self-service terminal integration, then studies the voice interaction principle based on artificial intelligence technology, and designs the integrated structure of financial self-service terminal with voice interaction. After that, this paper makes a series of tests on voice interaction technology, user experience, and the performance of financial self-service terminal. Finally, the test results of voice interaction are as follows: the delay estimation results of voice interaction of the terminal are relatively accurate, and the error points are basically within five sampling points, which indicate that the delay estimation algorithm is practical. The endpoint detection method based on CO complexity can effectively overcome the impact of noise environment on speech endpoint detection system and is suitable for the requirements of robust speech recognition system. Considering that the actual application scenario of voice positioning can judge the speaker’s position and turn to the speaker’s direction during human–computer interaction, the azimuth error is acceptable within a few degrees to meet the application requirements. The direction angle error is acceptable within a few degrees to meet the application requirements. The accuracy of the improved algorithm is improved in intercepting effective speech signals. The terminal has short running time and delay time, small memory, and central processing unit occupation and can meet the needs of users. The speech recognition accuracy of the financial self-service terminal basically reaches more than 80%, which can basically meet the daily needs. (shrink)

Top-down feedback does not benefit speech recognition; on the contrary, it can hinder it. No experimental data imply that feedback loops are required for speech recognition. Feedback is accordingly unnecessary and spoken word recognition is modular. To defend this thesis, we analyse lexical involvement in phonemic decision making. TRACE (McClelland & Elman 1986), a model with feedback from the lexicon to prelexical processes, is unable to account for all the available data on phonemic decision making. (...) The modular Race model (Cutler & Norris 1979) is likewise challenged by some recent results, however. We therefore present a new modular model of phonemic decision making, the Merge model. In Merge, information flows from prelexical processes to the lexicon without feedback. Because phonemic decisions are based on the merging of prelexical and lexical information, Merge correctly predicts lexical involvement in phonemic decisions in both words and nonwords. Computer simulations show how Merge is able to account for the data through a process of competition between lexical hypotheses. We discuss the issue of feedback in other areas of language processing and conclude that modular models are particularly well suited to the problems and constraints of speech recognition. Key Words: computational modeling; feedback; lexical processing; modularity; phonemic decisions; reading; speech recognition; word recognition. (shrink)

In this study, multidimensional feature extraction is performed on the U-language recordings of the test takers, and these features are evaluated separately, with five categories of features: pronunciation, fluency, vocabulary, grammar, and semantics. A deep neural network model is constructed to model the feature values to obtain the final score. Based on the previous research, this study uses a deep neural network training model instead of linear regression to improve the correlation between model score and expert score. The method of (...) using word frequency for semantic scoring is replaced by the LDA topic model for semantic analysis, which eliminates the need for experts to manually label keywords before scoring and truly automates the critique. Also, this paper introduces text cleaning after speech recognition and deep learning-based speech noise reduction technology in the scoring model, which improves the accuracy of speech recognition and the overall accuracy of the scoring model. Also, innovative applications and improvements are made to key technologies, and the latest technical solutions are integrated and improved. A new open oral grading model is proposed and implemented, and innovations are made in the method of speech feature extraction to improve the dimensionality of open oral grading. (shrink)

Through the recognition of ocean sediment sonar images, the texture in the image can be classified, which provides an important basis for the classification of ocean sediment. Aiming at the problems of low efficiency, waste of human resources, and low accuracy in the traditional manual side-scan sonar image discrimination, this paper studies the application of image recognition technology in sonar image substrate texture discrimination, which is popular in many fields. At the same time, considering the scale complexity, (...) diversity, multisources, and small sample characteristics of the marine sediment sonar image texture, the transfer learning is introduced into the image recognition, and the K-means clustering algorithm is used to reset the prior frame parameters to improve the speed and accuracy of image recognition. Through the experimental comparison between the original model and the new model based on transfer learning, the AP value of the yolov3 model based on transfer learning can reach 84.39%, which is 0.97% higher than that of the original model, with considerable accuracy and room for improvement; it takes less than 0.2 seconds. This shows the applicability and development of image recognition technology in texture discrimination of bottom sonar images. (shrink)

Speech recognition by second language (L2) learners in optimal and suboptimal conditions has been examined extensively with English as the target language in most previous studies. This study extended existing experimental protocols ( Wang et al., 2013 ) to investigate Mandarin speech recognition by Japanese learners of Mandarin at two different levels (elementary vs. intermediate) of proficiency. The overall results showed that in addition to L2 proficiency, semantic context, F0 contours, and listening condition all affected the (...) recognition performance on the Mandarin sentences. However, the effects of semantic context and F0 contours on L2 speech recognition diverged to some extent. Specifically, there was significant modulation effect of listening condition on semantic context, indicating that L2 learners made use of semantic context less efficiently in the interfering background than in quiet. In contrast, no significant modulation effect of listening condition on F0 contours was found. Furthermore, there was significant interaction between semantic context and F0 contours, indicating that semantic context becomes more important for L2 speech recognition when F0 information is degraded. None of these effects were found to be modulated by L2 proficiency. The discrepancy in the effects of semantic context and F0 contours on L2 speech recognition in the interfering background might be related to differences in processing capacities required by the two types of information in adverse listening conditions. (shrink)

Emotion recognition technology uses artificial intelligence to make inferences about a person’s emotions, on the basis of their facial expressions, body language, tone of voice, or other types of input. Underlying such technology are a variety of assumptions about the manifestation, nature, and value of emotions. To assure the quality and desirability of emotion recognition technology, it is important to critically assess the assumptions embedded in the technology. Within philosophy, there is a long tradition (...) of epistemological, ontological, phenomenological, and ethical reflection on the manifestation, nature, and value of emotions. This article draws from this tradition of philosophy of emotions, in order to challenge the assumptions underlying current emotion recognition technology and to promote a more critical engagement with the concept of emotions in the tech-industry. (shrink)

Hidden Markov models (HMMs) have been successful for modelling the dynamics of carefully dictated speech, but their performance degrades severely when used to model conversational speech. Since speech is produced by a system of loosely coupled articulators, stochastic models explicitly representing this parallelism may have advantages for automatic speech recognition (ASR), particularly when trying to model the phonological effects inherent in casual spontaneous speech. This paper presents a preliminary feasibility study of one such model (...) class: loosely coupled HMMs. Exact model estimation and decoding is potentially expensive, so possible approximate algorithms are also discussed. Comparison of one particular loosely coupled model on an isolated word task suggests loosely coupled HMMs merit further investigation. An approximate algorithm giving performance which is almost always statistically indistinguishable from the exact algorithm is also identified, making more extensive research computationally feasible. (shrink)

This paper implements the continuous Hindi Automatic Speech Recognition (ASR) system using the proposed integrated features vector with Recurrent Neural Network (RNN) based Language Modeling (LM). The proposed system also implements the speaker adaptation using Maximum-Likelihood Linear Regression (MLLR) and Constrained Maximum likelihood Linear Regression (C-MLLR). This system is discriminatively trained by Maximum Mutual Information (MMI) and Minimum Phone Error (MPE) techniques with 256 Gaussian mixture per Hidden Markov Model(HMM) state. The training of the baseline system has been (...) done using a phonetically rich Hindi dataset. The results show that discriminative training enhances the baseline system performance by up to 3%. Further improvement of ~7% has been recorded by applying RNN LM. The proposed Hindi ASR system shows significant performance improvement over other current state-of-the-art techniques. (shrink)

Despite the lack of invariance problem (the many‐to‐many mapping between acoustics and percepts), human listeners experience phonetic constancy and typically perceive what a speaker intends. Most models of human speech recognition (HSR) have side‐stepped this problem, working with abstract, idealized inputs and deferring the challenge of working with real speech. In contrast, carefully engineered deep learning networks allow robust, real‐world automatic speech recognition (ASR). However, the complexities of deep learning architectures and training regimens make it (...) difficult to use them to provide direct insights into mechanisms that may support HSR. In this brief article, we report preliminary results from a two‐layer network that borrows one element from ASR, long short‐term memory nodes, which provide dynamic memory for a range of temporal spans. This allows the model to learn to map real speech from multiple talkers to semantic targets with high accuracy, with human‐like timecourse of lexical access and phonological competition. Internal representations emerge that resemble phonetically organized responses in human superior temporal gyrus, suggesting that the model develops a distributed phonological code despite no explicit training on phonetic or phonemic targets. The ability to work with real speech is a major advance for cognitive models of HSR. (shrink)

This paper presents a novel philosophical analysis of the problem of law enforcement’s use of biased face recognition technology in liberal democracies. FRT programs used by law enforcement in identifying crime suspects are substantially more error-prone on facial images depicting darker skin tones and females as compared to facial images depicting Caucasian males. This bias can lead to citizens being wrongfully investigated by police along racial and gender lines. The author develops and defends “A Liberal Argument Against Biased (...) FRT,” which concludes that law enforcement use of biased FRT is inconsistent with the classical liberal requirement that government treat all citizens equally before the law. Two objections to this argument are considered and shown to be unsound. The author concludes by suggesting that equality before the law should be preserved while the problem of machine bias ought to be resolved before FRT and other types of artificial intelligence are deployed by governments in liberal democracies. (shrink)

Purpose This study aims to explore whether face recognition technology – as it is intensely used by state and local police departments and law enforcement agencies – is racism free or, on the contrary, is affected by racial biases and/or racist prejudices, thus reinforcing overall racial discrimination. Design/methodology/approach The study investigates the causal pathways through which face recognition technology may reinforce the racial disproportion in enforcement; it also inquires whether it further discriminates black people by making (...) them experience more racial discrimination and self-identify more decisively as black – two conditions that are shown to be harmful in various respects. Findings This study shows that face recognition technology, as it is produced, implemented and used in Western societies, reinforces existing racial disparities in stop, investigation, arrest and incarceration rates because of racist prejudices and even contributes to strengthen the unhealthy effects of racism on historically disadvantaged racial groups, like black people. Practical implications The findings hope to make law enforcement agencies and software companies aware that they must take adequate action against the racially discriminative effects of the use of face recognition technology. Social implications This study highlights that no implementation of an allegedly racism-free biometric technology is safe from the risk of racially discriminating, simply because each implementation leans against our society, which is affected by racism in many persisting ways. Originality/value While the ethical survey of biometric technologies is traditionally framed in the discourse of universal rights, this study explores an issue that has not been deeply scrutinized so far, that is, how face recognition technology differently affects distinct racial groups and how it contributes to racial discrimination. (shrink)

Vocabulary size has been suggested as a useful measure of “verbal abilities” that correlates with speech recognition scores. Knowing more words is linked to better speech recognition. How vocabulary knowledge translates to general speech recognition mechanisms, how these mechanisms relate to offline speech recognition scores, and how they may be modulated by acoustical distortion or age, is less clear. Age-related differences in linguistic measures may predict age-related differences in speech recognition (...) in noise performance. We hypothesized that speech recognition performance can be predicted by the efficiency of lexical access, which refers to the speed with which a given word can be searched and accessed relative to the size of the mental lexicon. We tested speech recognition in a clinical German sentence-in-noise test at two signal-to-noise ratios (SNRs), in 22 younger (18–35 years) and 22 older (60–78 years) listeners with normal hearing. We also assessed receptive vocabulary, lexical access time, verbal working memory, and hearing thresholds as measures of individual differences. Age group, SNR level, vocabulary size, and lexical access time were significant predictors of individual speech recognition scores, but working memory and hearing threshold were not. Interestingly, longer accessing times were correlated with better speech recognition scores. Hierarchical regression models for each subset of age group and SNR showed very similar patterns: the combination of vocabulary size and lexical access time contributed most to speech recognition performance; only for the younger group at the better SNR (yielding about 85% correct speech recognition) did vocabulary size alone predict performance. Our data suggest that successful speech recognition in noise is mainly modulated by the efficiency of lexical access. This suggests that older adults’ poorer performance in the speech recognition task may have arisen from reduced efficiency in lexical access; with an average vocabulary size similar to that of younger adults, they were still slower in lexical access. (shrink)

Norris, McQueen & Cutler claim that all known speech recognition data can be accounted for with their autonomous model, “Merge.” But this claim is doubly misleading. (1) Although speech recognition is autonomous in their view, the Merge model is not. (2) The body of data which the Merge model accounts for, is not, in their view, speech recognition data. Footnotes1 Author is also affiliated with the Center for the Study of Language and Information, Stanford (...) University, Stanford, CA 94305, [email protected]. (shrink)

This paper combines domestic and international research results to analyze and study the difference between the attribute features of English phrase speech and noise to enhance the short-time energy, which is used to improve the threshold judgment sensitivity; noise addition to the discrepancy data set is used to enhance the recognition robustness. The backpropagation algorithm is improved to constrain the range of weight variation, avoid oscillation phenomenon, and shorten the training time. In the real English phrase sound (...) class='Hi'>recognition system, there are problems such as massive training data and low training efficiency caused by the super large-scale model parameters of the convolutional neural network. To address these problems, the NWBP algorithm is based on the oscillation phenomenon that tends to occur when searching for the minimum error value in the late training period of the network parameters, using the K-MEANS algorithm to obtain the seed nodes that approach the minimal error value, and using the boundary value rule to reduce the range of weight change to reduce the oscillation phenomenon so that the network error converges as soon as possible and improve the training efficiency. Through simulation experiments, the NWBP algorithm improves the degree of fitting and convergence speed in the training of complex convolutional neural networks compared with other algorithms, reduces the redundant computation, and shortens the training time to a certain extent, and the algorithm has the advantage of accelerating the convergence of the network compared with simple networks. The word tree constraint and its efficient storage structure are introduced, which improves the storage efficiency of the word tree constraint and the retrieval efficiency in the English phrase recognition search. (shrink)

Previous work demonstrated that dynamic pitch (i.e., pitch variation in speech) aids speech recognition in various types of noises. While this finding suggests dynamic pitch enhancement in target speech can benefit speech recognition in noise, it is of importance to know what noise characteristics affect dynamic pitch benefit and who will benefit from enhanced dynamic pitch cues. Following our recent finding that temporal modulation in noise influences dynamic pitch benefit, we examined the effect of (...) speech masker characteristics on dynamic pitch benefit. Specifically, the first goal of the study was to test the hypothesis that dynamic pitch benefit varies depending on the availability of pitch cues in the masker and the intelligibility of masker. The second goal of this study was to investigate whether older listeners as a group can benefit from dynamic pitch for speech recognition in speech maskers. In addition, individual factors of hearing loss and working memory capacity were examined for their impact on older listeners’ dynamic pitch benefit. Twenty-three younger listeners with normal hearing and 37 older listeners with varying levels of hearing sensitivity participated the study, in which speech reception thresholds were measured with sentences in speech maskers. While we did not find an effect of masker characteristics on dynamic pitch benefit, the results showed older listeners can benefit from dynamic pitch for recognizing speech in speech maskers. The data also suggest that among those older listeners with hearing loss, dynamic pitch benefit is stronger for individuals with higher working memory capacity. This can be attributed to their ability to exploit facilitated lexical access in processing of degraded speech signal. (shrink)

In this paper I present the position that the use of face recognition technology (FRT) in law enforcement and in business is restrictive of individual autonomy. I reason that FRT severely undermines autonomous self-determination by hobbling the idea of freedom of the will. I distinguish this position from two other common arguments against surveillance technologies: the privacy argument (that FRT is an invasion of privacy) and the objective freedom argument (that FRT is restrictive of one's freedom to act). (...) To make this case, I suggest that autonomy itself is predicated on the possibility of acting ethically, of freely willing moral laws. I then claim that autonomous self-determination is established as self-determination via social interactions with others. If we conceptualize self-determination as a relation of establishing a claim to individual autonomy in a community of others, we can see how planned uses of FRT subvert possibilities for the establishment of socially recognized agency. FRT not only confuses the process of asking ethical questions but it also imposes the immanent likelihood that all actions are taken not by self-directed, free agents, but by passive subjects in the interest of abiding by the institutionally enforced law. (shrink)

Deep neural networks (DNNs) have been playing a significant role in acoustic modeling. Convolutional neural networks (CNNs) are the advanced version of DNNs that achieve 4–12% relative gain in the word error rate (WER) over DNNs. Existence of spectral variations and local correlations in speech signal makes CNNs more capable of speech recognition. Recently, it has been demonstrated that bidirectional long short-term memory (BLSTM) produces higher recognition rate in acoustic modeling because they are adequate to reinforce (...) higher-level representations of acoustic data. Spatial and temporal properties of the speech signal are essential for high recognition rate, so the concept of combining two different networks came into mind. In this paper, a hybrid architecture of CNN-BLSTM is proposed to appropriately use these properties and to improve the continuous speech recognition task. Further, we explore different methods like weight sharing, the appropriate number of hidden units, and ideal pooling strategy for CNN to achieve a high recognition rate. Specifically, the focus is also on how many BLSTM layers are effective. This paper also attempts to overcome another shortcoming of CNN, i.e. speaker-adapted features, which are not possible to be directly modeled in CNN. Next, various non-linearities with or without dropout are analyzed for speech tasks. Experiments indicate that proposed hybrid architecture with speaker-adapted features and maxout non-linearity with dropout idea shows 5.8% and 10% relative decrease in WER over the CNN and DNN systems, respectively. (shrink)

Biometric facial recognition is an artificial intelligence technology involving the automated comparison of facial features, used by law enforcement to identify unknown suspects from photographs and closed circuit television. Its capability is expanding rapidly in association with artificial intelligence and has great potential to solve crime. However, it also carries significant privacy and other ethical implications that require law and regulation. This article examines the rise of biometric facial recognition, current applications and legal developments, and conducts an (...) ethical analysis of the issues that arise. Ethical principles are applied to mediate the potential conflicts in relation to this information technology that arise between security, on the one hand, and individual privacy and autonomy, and democratic accountability, on the other. These can be used to support appropriate law and regulation for the technology as it continues to develop. (shrink)

This article presents an overview of the single-channel dereverberation methods suitable for distant speech recognition application. The dereverberation methods are mainly classified based on the domain of enhancement of speech signal captured by a distant microphone. Many single-channel speech enhancement methods focus on either denoising or dereverberating the distorted speech signal. There are very few methods that consider both noise and reverberation effects. Such methods are discussed under a multistage approach in this article. The article (...) concludes with a hypothesis that the methods that do not require an a priori reverberation impulse response is desirable in varying the environmental conditions for DSR applications such as intelligent home and office environments, humanoid robots, and automobiles rather than the methods that require an a priori reverberation impulse response. (shrink)

In this paper, we propose to incorporate the local attention in WaveNet-CTC to improve the performance of Tibetan speech recognition in multitask learning. With an increase in task number, such as simultaneous Tibetan speech content recognition, dialect identification, and speaker recognition, the accuracy rate of a single WaveNet-CTC decreases on speech recognition. Inspired by the attention mechanism, we introduce the local attention to automatically tune the weights of feature frames in a window and (...) pay different attention on context information for multitask learning. The experimental results show that our method improves the accuracies of speech recognition for all Tibetan dialects in three-task learning, compared with the baseline model. Furthermore, our method significantly improves the accuracy for low-resource dialect by 5.11% against the specific-dialect model. (shrink)

On February 13, 2020, the Toronto Police Services (TPS) issued a statement admitting that its members had used Clearview AI’s controversial facial recognition technology (FRT). The controversy sparked widespread outcry by the media, civil society, and community groups, and put pressure on policy-makers to address FRTs. Public consultations presented a key tool to contain the scandal in Toronto and across Canada. Drawing on media reports, policy documents, and expert interviews, we investigate four consultations held by the Toronto Police (...) Services Board (TPSB), the Office of the Privacy Commissioner (OPC), and the parliamentary Standing Committee on Access to Information, Privacy and Ethics (ETHI) to understand how public opinion and outrage translate into policy. We find that public consultations became a powerful closure mechanism in the policy-making toolbox, inhibiting rather than furthering democratic debate. Our findings show that consultations do not advance public literacy; that opportunities for public input are narrow; that timeframes are short; and that mechanisms for inclusion are limited. Even in the best-case circumstances, consultations are merely one of many factors in AI governance and seldom impact concrete policy outcomes in the cases studied here. (shrink)

This study examines how different forms of media use predict attitudes toward the development of facial recognition technology (FRT) and applications of it by law enforcement to identify criminal suspects, identify potential terrorists, and monitor public protests. The theoretical framework builds on theories of cultivation and genre-specific viewing to develop hypotheses and research questions. The analyses draw on original data from two nationally representative surveys of the U.S. public conducted in 2020, amid a series of controversies and protests (...) about policing and racial justice. The results demonstrate that overall television viewing and crime media viewing predicted support for multiple uses of FRT, while Fox News viewing predicted support for using FRT to monitor protests. The findings advance our understanding of public opinion toward the technology and its implications for policing, protests, and social justice. (shrink)

As more and more companies adopt artificial intelligence to increase the efficiency and effectiveness of their products and services, they expose themselves to ethical crises and potentially damaging public controversy associated with its use. Despite the prevalence of AI ethical problems, most companies are strategically unprepared to respond effectively to the public. This paper aims to advance our empirical understanding of company responses to AI ethical crises by focusing on the rise and fall of facial recognition technology. Specifically, (...) through a comparative case study of how four big technology companies responded to public outcry over their facial recognition programs, we not only demonstrated the unfolding and consequences of public controversies over this new technology, but also identified and described four major types of company responses—Deflection, Improvement, Validation, and Pre-emption. These findings pave the way for future research on the management of controversial technology and the ethics of AI. (shrink)

One aspect of auditory scenes that has received very little attention is the level of diffuseness of sound sources. This aspect has increasing importance due to growing use of amplification systems. When an auditory stimulus is amplified and presented over multiple, spatially-separated loudspeakers, the signal’s timbre is altered due to comb filtering. In a previous study we examined how increasing the diffuseness of the sound sources might affect listeners’ ability to recognize speech presented in different types of background noise. (...) Listeners performed similarly when both the target and the masker were presented via a similar number of loudspeakers. However, performance improved when the target was presented using a single speaker and the masker from three spatially separate speakers but worsened when the target was diffuse, and the masker was compact. In the current study, we extended our research to examine whether the effects of timbre changes with age and linguistic experience. Twenty-four older adults whose first language was English and 24 younger adults whose second language was English were asked to repeat non-sense sentences masked by either Noise, Babble, or Speech and their results were compared with those of the Young-EFLs previously tested. Participants were divided into two experimental groups: A Compact-Target group where the target sentences were presented over a single loudspeaker, while the masker was either presented over three loudspeakers or over a single loudspeaker; A Diffuse-Target group, where the target sentences were diffuse while the masker was either compact or diffuse. The results indicate that the Target Timbre has a negligible effect on thresholds when the timbre of the target matches the timbre of the masker in all three groups. When there is a timbre contrast between target and masker, thresholds are significantly lower when the target is compact than when it is diffuse for all three listening groups in a Noise background. However, while this difference is maintained for the Young and Old-EFLs when the masker is Babble or Speech, speech reception thresholds in the Young-ESL group tend to be equivalent for all four combinations of target and masker timbre. (shrink)