Speech Classification Research Articles

Subject-independent meta-learning framework towards optimal training of EEG-based classifiers

Advances in deep learning have shown great promise towards the application of performing high-accuracy Electroencephalography (EEG) signal classification in a variety of tasks. However, many EEG-based datasets are often plagued by the issue of high inter-subject signal variability. Robust deep learning models are notoriously difficult to train under such scenarios, often leading to subpar or widely varying performance across subjects under the leave-one-subject-out paradigm. Recently, the model agnostic meta-learning framework was introduced as a way to increase the model’s ability to generalize towards new tasks. While the original framework focused on task-based meta-learning, this research aims to show that the meta-learning methodology can be modified towards subject-based signal classification while maintaining the same task objectives and achieve state-of-the-art performance. Namely, we propose the novel implementation of a few/zero-shot subject-independent meta-learning framework towards multi-class inner speech and binary class motor imagery classification. Compared to current subject-adaptive methods which utilize large number of labels from the target, the proposed framework shows its effectiveness in training zero-calibration and few-shot models for subject-independent EEG classification. The proposed few/zero-shot subject-independent meta-learning mechanism performs well on both small and large datasets and achieves robust, generalized performance across subjects. The results obtained shows a significant improvement over the current state-of-the-art, with the binary class motor imagery achieving 88.70% and the accuracy of multi-class inner speech achieving an average of 31.15%. Codes will be made available to public upon publication.

Significance of relative phase features for shouted and normal speech classification

Shouted and normal speech classification plays an important role in many speech-related applications. The existing works are often based on magnitude-based features and ignore phase-based features, which are directly related to magnitude information. In this paper, the importance of phase-based features is explored for the detection of shouted speech. The novel contributions of this work are as follows. (1) Three phase-based features, namely, relative phase (RP), linear prediction analysis estimated speech-based RP (LPAES-RP) and linear prediction residual-based RP (LPR-RP) features, are explored for shouted and normal speech classification. (2) We propose a new RP feature, called the glottal source-based RP (GRP) feature. The main idea of the proposed GRP feature is to exploit the difference between RP and LPAES-RP features to detect shouted speech. (3) A score combination of phase- and magnitude-based features is also employed to further improve the classification performance. The proposed feature and combination are evaluated using the shouted normal electroglottograph speech (SNE-Speech) corpus. The experimental findings show that the RP, LPAES-RP, and LPR-RP features provide promising results for the detection of shouted speech. We also find that the proposed GRP feature can provide better results than those of the standard mel-frequency cepstral coefficient (MFCC) feature. Moreover, compared to using individual features, the score combination of the MFCC and RP/LPAES-RP/LPR-RP/GRP features yields an improved detection performance. Performance analysis under noisy environments shows that the score combination of the MFCC and the RP/LPAES-RP/LPR-RP features gives more robust classification. These outcomes show the importance of RP features in distinguishing shouted speech from normal speech.

AN AUTOMATIC HATE SPEECH DETECTION IN SOCIAL MEDIA THROUGH COMPUTATIONAL LINGUISTICS: INFIDELITY VIDEOS IN FOCUS

The escalating prevalence of hate speeches, amplified by the misguided use of social media, introduces alarming challenges to the safeguarding of human rights and individual welfare. Motivated by this, the study explored the detection and classification of hate speech, specifically as observed in speeches and comments related to infidelity videos on YouTube Channel of Raffy Tulfo in Action. Further, the study utilized a computational linguistic algorithm through Long Short-Term Memory (LSTM). Additionally, the study sought to understand the distinctions in linguistic features between hate speech and non-hateful speech through LSTM. The researcher used 9,600,586 tokens for the analysis. To answer the first research question, the employment of LSTM helped identify hate speeches from non-hate speeches through effective data gathering through YouTube Application Programming Interface (API) and Whisper AI, text processing, labeling, coding, and algorithm deployment. Through that process, LSTM also classified them per target, including sex, quality, physical attributes, disability, religion, race, and class. Further, to answer the second research question, the study was able to identify 35 lexicons. Some samples include peenoise, U10, kokey, taitok, quibolok, squami, and shut@, which were used negatively. Lastly, to answer the last question, tokenization, embedding, sequential dependencies, padding, training-testing, and evaluating helped LSTM assess hate speech linguistic features. It is evident in the confusion matrix showing 46% true positives and 49% true negatives and its evaluation performance of 95% F1 score, affirming its high robustness and reliability. KEYWORDS: Applied linguistics, language, hates speeches, infidelity cases, computational linguistics, Long Short Term-Memory (LSTM), Philippines

Dual-Mode Method for Generating Adversarial Examples to Attack Deep Neural Networks

Deep neural networks yield desirable performance in text, image, and speech classification. However, these networks are vulnerable to adversarial examples. An adversarial example is a sample generated by inserting a small amount of noise into an original sample (with minimal distortion) such that it is recognized incorrectly by the targeted model. A typical method of attack using adversarial examples must satisfy two conditions: the distortion of the original sample must be kept to a minimum and misrecognition must be induced in the targeted deep neural network. Therefore, considerable time and numerous iterations are required to generate an adversarial example because both the conditions must be satisfied during the generation process. However, there are cases in which it may be desirable to generate an adversarial example that acts quickly to induce misrecognition in the deep neural network without considering the amount of distortion applied to the original sample. In this paper, we propose a dual-mode method for creating adversarial examples that allows the user to prioritize the malfunctioning of deep neural networks according to the situation. The proposed method generates an adversarial example using one of two modes: mode 1, which takes the level of distortion into account, and mode 0, which does not consider distortion and can generate examples rapidly. To evaluate the method experimentally, MNIST and CIFAR10 were used as the datasets. The results show that the proposed method can generate a targeted or untargeted adversarial example for MNIST with 50% fewer iterations using mode 0 than using mode 1. For CIFAR10, the reduction in the number of iterations that can be achieved using mode 0 is 80% and 88% for targeted and untargeted adversarial examples, respectively, and the attack success rate is 100%.

The classifications of artistic speech – content, analysis and practical styles of music art

The classifications of artistic speech – content, analysis and practical styles of music art

Deep Learning-Based Feature Extraction for Speech Emotion Recognition

Emotion recognition from speech signals is an important and challenging component of Human-Computer Interaction. In the field of speech emotion recognition (SER), many techniques have been utilized to extract emotions from speech signals, including many well-established speech analysis and classification techniques. This model can be built by using various methods such as RNN, SVM, deep learning, cepstral coefficients, and various other methods, out of which SVM normally gives us the highest accuracy. We propose a model that can identify emotions present in the speech, which can be identified by various parameters such as pitch, speaking rate, speech time, and frequency patterns. Emotion detection in digitized speech contains 3 components: Signal processing, Feature extraction, and Classification. The model first tries to remove the background noises then extract the features present in the speech and classify it into a single emotion. This model is capable of identifying seven different emotions that can be found in human speech. We can use different classifiers like GMM and HMM to classify features such as Spectral Subtraction, Wiener Filtering, Adaptive Filtering, and Deep Learning Techniques. This model can be used in various fields such as healthcare, security, psychology, medicine, education, and entertainment.

Hybrid Transformer Architectures with Diverse Audio Features for Deepfake Speech Classification

Hybrid Transformer Architectures with Diverse Audio Features for Deepfake Speech Classification

A BCI System for Imagined Speech Classification Based on Optimization Theory

A BCI System for Imagined Speech Classification Based on Optimization Theory

Hate speech detection with ADHAR: a multi-dialectal hate speech corpus in Arabic.

Hate speech detection in Arabic poses a complex challenge due to the dialectal diversity across the Arab world. Most existing hate speech datasets for Arabic cover only one dialect or one hate speech category. They also lack balance across dialects, topics, and hate/non-hate classes. In this paper, we address this gap by presenting ADHAR-a comprehensive multi-dialect, multi-category hate speech corpus for Arabic. ADHAR contains 70,369 words and spans four language variants: Modern Standard Arabic (MSA), Egyptian, Levantine, Gulf and Maghrebi. It covers four key hate speech categories: nationality, religion, ethnicity, and race. A major contribution is that ADHAR is carefully curated to maintain balance across dialects, categories, and hate/non-hate classes to enable unbiased dataset evaluation. We describe the systematic data collection methodology, followed by a rigorous annotation process involving multiple annotators per dialect. Extensive qualitative and quantitative analyses demonstrate the quality and usefulness of ADHAR. Our experiments with various classical and deep learning models demonstrate that our dataset enables the development of robust hate speech classifiers for Arabic, achieving accuracy and F1-scores of up to 90% for hate speech detection and up to 92% for category detection. When trained with Arabert, we achieved an accuracy and F1-score of 94% for hate speech detection, as well as 95% for the category detection.

Voices on the air: Speech education and campus radio in the postcolonial Philippine university

In the late 1950s and throughout the 1960s, faculty and students of the newly organized Department of Speech and Drama (later Department of Speech Communication and Theatre Arts) at the University of the Philippines (UP) were at the forefront of managing the radio station DZUP, mounting radio productions on campus and shaping the academic curricula for classes in radio speech and writing. These pioneering contributions, though significant, have yet to receive due documentation from communication scholars, researchers, and historians in the country. In this essay, I address this gap by bringing into focus archival documents—photographs, newspaper accounts, official memos, and personal correspondences between academics and administrators—that clarify the academic department’s role in the early systems and operations of DZUP. I argue that these efforts are important because of four main reasons. First, they highlight the often-overlooked relationship between speech education and campus radio in the national university. Second, they emphasize the ways in which the radio booth worked alongside the public speaking platform and the theatre stage as a fundamental space where speech-related pedagogies, performances, and practices played out. Third, they show disciplinary genealogy that links the disciplines of speech communication and mass communication in the University of the Philippines. And finally, they shed light on the pedagogical process involved in teaching, training, and transforming Filipino students into a kind of speaking subjects in the postcolonial Philippines.

DIACRITIC-AWARE ALIGNMENT AND CLASSIFICATION IN ARABIC SPEECH: A FUSION OF FUZTPI AND ML MODELS

This paper presents the Quran Speech Recognition (QRSR) system, achieving alignment and classification accuracies up to 96%. The system is designed to advance Arabic Automatic Speech Recognition (ASR) and Text-to-Speech (TTS) by focusing on the Arabic diacritic-annotated text. We address the limitations of existing Arabic ASR systems and introduce the Fuzzy Text Alignment and Rule-based Classifier (FTARC) for segmenting audio files and aligning text. The FuzTPI algorithm is integrated with Machine Learning models like Na¨ıve Bayes, Support Vector Machine, and Random Forest. This research aims to generalize the findings for broader Arabic text and contribute to an expanded audio dataset, thereby enhancing Arabic NLP and speech recognition capabilities.

화행중심 교수학습전략 사용이 영어 학습자의 화행 습득에 끼치는 영향

The purpose of this study is to investigate current and classical research on speech acts and indirect speech acts, determine how to teach them to English learners to improve their English proficiency and provide how to implement teaching strategies related to speech acts in the EFL classroom. Section II presents Austin’s (1962) definition and classification of speech acts. In Section III, Searle’s (1969) theory of speech acts is explored, and other relevant topics for the language classroom are reviewed. In Section IV, the differences between these and antitheses are presented, providing some reasons for the collapse of the antitheses. Finally, concluding remarks on the characteristics of speech and indirect speech acts are presented, providing teaching strategies for Korean EFL classrooms.

An Analysis of Speech Acts in Caprio’s Speech in American First Instance Court Pleadings

The current study deals with the analysis of speech acts utilized by Judge Frank Caprio in American first instance court pleading. The study attempts to determine the types of speech acts used in the selected data by the judge Caprio. It also aims to is to identify the most frequent major functions of illocutionary acts, as well as determining the level of directness (direct or indirect speech). Searle’s (1976) classification of speech acts is used as a model to analyze the utterances. The results indicate that the judge Caprio employs the four types of speech acts based on Searle’s theory, including directive, assertive, expressive, and declarative. Asking and stating are the most frequent functions used in the judge Caprio’s speech. Finally, the study concludes that direct speech act represents the majority of Judge Caprio's speech which indicates the high value of clarity in his utterances.

Analisis DistilBERT dengan Support Vector Machine (SVM) untuk Klasifikasi Ujaran Kebencian pada Sosial Media Twitter

Hate speech is a significant issue in content management on social media platforms. Effective classification of hate speech plays a crucial role in maintaining a safe social media environment, combating discrimination, and protecting users. This study evaluates a hate speech classification model using SVM with linear and polynomial kernels. The dataset used consists of labeled Indonesian-language tweets. The importance of developing an effective classification model to address hate speech has led to the utilization of DistilBERT as a feature extraction method. However, DistilBERT has high-dimensional features, necessitating dimensionality reduction to reduce model complexity. Therefore, in this study, the PCA dimensionality reduction method is implemented with various scenarios of dimensionality, namely 10, 20, 30, 40, and 50. Evaluation is performed using F1-Score, and the entire study is evaluated using 10-fold cross-validation. The evaluation results indicate that in the scenario with a linear kernel, the model achieves the highest F1-Score of 0.75 in the 50-dimensional scenario. Meanwhile, in the scenario with a polynomial kernel, the model achieves the highest F1-Score of 0.7857 in the 50-dimensional scenario. These findings demonstrate that the use of a polynomial kernel with 50 dimensions yields the best performance in classifying hate speech.

ILLOCUTIONARY SPEECH ACTS IN THE PRESIDENTIAL SPEECH REGARDING THE RELOCATION OF THE NATIONAL CAPITAL CITY

This study aims to analyze illocutionary speech acts in speeches on the relocation of the national capital. The relocation of the national capital is a new polemic that raises pros and cons among the public. Some people consider the relocation of the capital city to have less urgency, so there is no need to do that. Leaders have an important role in convincing the public of this view. The use of language by politicians in speeches certainly has another goal, namely to convince the audience or the public. This research uses the theory of speech acts by Searle. Searle classifies illocutionary speech acts into five types. That is a declaration, assertive, expressive, directive, and commissive. The method used is the descriptive method. The results obtained from this study are that there are classifications of speech acts. That is, assertive, commissive, directive, and declarative. The results of this study indicate that these speech acts have a role in speech, one of which is to influence many people with our opinions.

Do users want platform moderation or individual control? Examining the role of third-person effects and free speech support in shaping moderation preferences

This study examines social media users’ preferences for the use of platform-wide moderation in comparison to user-controlled, personalized moderation tools to regulate three categories of norm-violating content—hate speech, sexually explicit content, and violent content. Via a nationally representative survey of 984 US adults, we explore the influence of third-person effects and support for freedom of expression on this choice. We find that perceived negative effects on others negatively predict while free speech support positively predicts a preference for having personal moderation settings over platform-directed moderation for regulating each speech category. Our findings show that platform governance initiatives need to account for both actual and perceived media effects of norm-violating speech categories to increase user satisfaction. Our analysis also suggests that users do not view personal moderation tools as an infringement on others’ free speech but as a means to assert greater agency over their social media feeds.

Machine learning in healthcare strategic management: a systematic literature review

PurposeThis study specifically seeks to investigate the strategic implementation of machine learning (ML) algorithms and techniques in healthcare institutions to enhance innovation management in healthcare settings.Design/methodology/approach The papers from 2011 to 2021 were considered following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. First, relevant keywords were identified, and screening was performed. Bibliometric analysis was performed. One hundred twenty-three relevant documents that passed the eligibility criteria were finalized.Findings Overall, the annual scientific production section results reveal that ML in the healthcare sector is growing significantly. Performing bibliometric analysis has helped find unexplored areas; understand the trend of scientific publication; and categorize topics based on emerging, trending and essential. The paper discovers the influential authors, sources, countries and ML and healthcare management keywords.Research limitations/implications The study helps understand various applications of ML in healthcare institutions, such as the use of Internet of Things in healthcare, the prediction of disease, finding the seriousness of a case, natural language processing, speech and language-based classification, etc. This analysis would help future researchers and developers target the healthcare sector areas that are likely to grow in the coming future.Practical implications The study highlights the potential for ML to enhance medical support within healthcare institutions. It suggests that regression algorithms are particularly promising for this purpose. Hospital management can leverage time series ML algorithms to estimate the number of incoming patients, thus increasing hospital availability and optimizing resource allocation. ML has been instrumental in the development of these systems. By embracing telemedicine and remote monitoring, healthcare management can facilitate the creation of online patient surveillance and monitoring systems, allowing for early medical intervention and ultimately improving the efficiency and effectiveness of medical services.Originality/value By offering a comprehensive panorama of ML's integration within healthcare institutions, this study underscores the pivotal role of innovation management in healthcare. The findings contribute to a holistic understanding of ML's applications in healthcare and emphasize their potential to transform and optimize healthcare delivery.

Examining Speech Acts in Jordanian Advertising: Pragmatic Functions, Linguistic Features, and Rhetorical Devices

This qualitative study aims to investigate the persuasive speech acts used by Jordanian advertisers in fashion, food, and technology print advertisements. The selected advertisements were transcribed and analyzed using NVivo. Commercial speech acts were classified using Searle’s (1979) taxonomy of assertives, directives, commissives, expressives, and declaratives. The researcher took care to obtain the participants' informed consent and kept subject data confidential and anonymous. The evaluation of Jordanian advertising speech activities and the linguistic and cultural influences exerted upon them was conducted using tables and narratives. The findings of this research emphasize the diverse advertising strategies employed by Jordanian organizations. It was determined that directives, statements, and demands were the three most prevalent categories of speech acts in Jordanian advertising. These methods of communication were utilized to advocate for particular client actions, to make claims about the quality of the product being sold, and to guarantee certain outcomes for buyers. The study also revealed that Jordanian businesses frequently employ hyperbole, metaphor, and rhetorical queries in their marketing communications in order to make their products or services appear more desirable to potential consumers.

The Role of Consumer Speech Acts in Brand Activism: A Transformative Advertising Perspective

Transformative advertising research (TAR) suggests examining advertising’s transformational possibilities via the interactions between institutional actors at each marketing level to gauge its effect on society. We employ rhetorical institutionalism as a lens to examine the online speech acts of consumers as they respond to a brand activism campaign focusing on an environmental problem. Our data take the form of written comments by YouTube users and employ a research design using automated text analysis and qualitative thematic data analysis. Our contributions to TAR are threefold. First, we offer a preliminary conceptualization of the role of consumer language as rhetorical institutional work to advance TAR scholars and practitioners’ insight. Second, we highlight the role of linguistic tone and clout in giving speakers agency through which consumers as institutional actors create, maintain, and disrupt institutional logics and practices. Finally, we develop a tripartite classification of consumer speech acts used to support brand activism. We label these activist warriors, brand champions, and conscious consumers as typologies that deepen understanding of how consumers’ online speech may amplify brand activism, thereby contributing to advertising’s transformative outcomes. We conclude by outlining important managerial implications including how practitioners can adopt the tripartite classification to enhance brand activism campaigns.

Ecologically valid speech collection in behavioral research: The Ghent Semi-spontaneous Speech Paradigm (GSSP)

This paper introduces the Ghent Semi-spontaneous Speech Paradigm (GSSP), a new method for collecting unscripted speech data for affective-behavioral research in both experimental and real-world settings through the description of peer-rated pictures with a consistent affective load. The GSSP was designed to meet five criteria: (1) allow flexible speech recording durations, (2) provide a straightforward and non-interfering task, (3) allow for experimental control, (4) favor spontaneous speech for its prosodic richness, and (5) require minimal human interference to enable scalability. The validity of the GSSP was evaluated through an online task, in which this paradigm was implemented alongside a fixed-text read-aloud task. The results indicate that participants were able to describe images with an adequate duration, and acoustic analysis demonstrated a trend for most features in line with the targeted speech styles (i.e., unscripted spontaneous speech versus scripted read-aloud speech). A speech style classification model using acoustic features achieved a balanced accuracy of 83% on within-dataset validation, indicating separability between the GSSP and read-aloud speech task. Furthermore, when validating this model on an external dataset that contains interview and read-aloud speech, a balanced accuracy score of 70% is obtained, indicating an acoustic correspondence between the GSSP speech and spontaneous interviewee speech. The GSSP is of special interest for behavioral and speech researchers looking to capture spontaneous speech, both in longitudinal ambulatory behavioral studies and laboratory studies. To facilitate future research on speech styles, acoustics, and affective states, the task implementation code, the collected dataset, and analysis notebooks are available.