Text-to-speech Research Articles

Tamil Speech Synthesizer App for Android: Text Processing Module Enhancement

Objectives: Designing dynamic computer systems that are effective, efficient, simple, and satisfying to use is becoming extremely important in this age of information and communication technology. Text to Speech or Speech Synthesis is one of the many methods being investigated by researchers to improve Human-Computer Interaction. The goal here is to improve the text processing component of the Tamil voice synthesizer by including a text normalizer and loan word identification that is efficient and reliable. Methods: Text normalization is conducted on unconstrained Tamil text to turn non-standard terms into common words to reduce confusing utterances during intermediate word processing. Loan/Native words in Tamil literature are detected to enhance the Tamil voice synthesizer system’s pronunciation model. Findings: During normalization, non-standard Tamil words are replaced with standard ones to reduce ambiguous utterances during interim processing. A pronunciation model is built to improve the Tamil speech synthesizer system by identifying loan words in Tamil text. A syllable classifier is presented in this study, based on a decision list approach, which can handle various types of non-stationary sounds. Novelty: We also disclose a ’loan/native word classifier’ based on multiple linear regressions that perform well even with small words of three syllables. Such sophisticated text processors are required in today’s dominating Digital, Information-Communication Technology, and Human-Computer Interaction age. Keywords: Mobile Communication Technology; HumanComputer Interaction; Speech Synthesis Affirms; Syllable Classifier; Prerecorded database

LimitAccess: on-device TinyML based robust speech recognition and age classification

Automakers from Honda to Lamborghini are incorporating voice interaction technology into their vehicles to improve the user experience and offer value-added services. Speech recognition systems are a key component of smart cars, enhancing convenience and safety for drivers and passengers. In the future, safety-critical features may rely on speech recognition, but this raises concerns about children accessing such services. To address this issue, the LimitAccess system is proposed, which uses TinyML for age classification and helps parents limit children’s access to critical speech recognition services. This study employs a lite convolutional neural network (CNN) model for two different reasons: First, CNN showed superior accuracy compared to other audio classification models for age classification problems. Second, the lite model will be integrated into a microcontroller to meet its limited resource requirements. To train and evaluate our model, we created a dataset that included child and adult voices of the keyword “open”. The system approach categorizes voices into age groups (child, adult) and then utilizes that categorization to grant access to a car. The robustness of the model was enhanced by adding a new class (recordings) to the dataset, which enabled our system to detect replay and synthetic voice attacks. If an adult voice is detected, access to start the car will be granted. However, if a child’s voice or a recording is detected, the system will display a warning message that educates the child about the dangers and consequences of the improper use of a car. Arduino Nano 33 BLE sensing was our embedded device of choice for integrating our trained, optimized model. Our system achieved an overall F1 score of 87.7% and 85.89% accuracy. LimitAccess detected replay and synthetic voice attacks with an 88% F1 score.

Blended teaching of English speaking in a Vietnamese context: Design and procedure

This article aims to disseminate a technology‐enhanced language pedagogy for teaching oral English in a World Englishes context, which is underpinned by the theories of scaffolding, scaffolded instruction, and task‐based learning, and utilises a Moodle learning management system (LMS) as its e‐learning platform. In this pedagogy, not only do technologies assist language teaching and learning in terms of lesson design, materials development, and lesson delivery, but they actually involve and play a crucial role in students’ oral language development with the exploitation of text‐to‐ speech (TTS) and automatic speech recognition (ASR). The contemporary status of teaching and learning oral English is discussed to clarify the congruence of the pedagogy with an emerging consensus on teaching English speaking. A brief summary of technology‐enhanced language learning (TELL) and a critical examination of current practices of teaching oral English with technology are also presented for consideration of the position of the proposed framework in the field. The adaptability of the framework in case of the inaccessibility of an LMS is also discussed.

Accessible Bar Charts Through Textual Description Templates

Data charts are very prevalent in everyday life in different contexts, from economics to politics. However, people with blindness and low vision do not have easy access to this information since they use screen reader software. This software does not extract the information available graphically, but only the chart legend and the text. Among the solutions proposed in the literature, there are crowdsourcing techniques when a person is responsible for interpreting the chart, which can cause bias in the chart’s interpretation. To solve this problem, we proposed textual description templates for simple and grouped bar charts to inform the chart data in a standardized way to users, excluding the interpretation bias. The methodology of this work was divided into three stages: a definition of templates for textual description and testing with 30 participants; the application of textual description templates in an assistive technology tool and testing with 45 participants; the validation of the results found through interviews and tests with 3 specialists. We have iteratively refined templates generated at each stage with users tests, and we carried out quantitative and qualitative analyses. An assistive technology tool, ChartVision, was developed to consume the templates. Finally, we interviewed a specialist about how he would explain chart materials to blind students at university, and we carried out a validation of the final templates with two other professionals from the health and education areas who deal with people with blind people in their daily lives. The main contributions are three textual description bar charts templates: simple bar for applications with sequential reading or reading on-demand, grouped bar for applications with sequential reading, and grouped bar for applications on-demand. The secondary contribution is ChartVision. Other findings include considerations about the synthetic voice used in the tests, expected characteristics for a better understanding of the chart, and interaction ways to access the information.

The synthetic green voice of tetrahydro-11H-benzo[a]xanthen-11-one derivatives by investigating the catalytic performance of the quadratic form of the organic–inorganic tetramethylammonium hexafluorosilicate [(CH3)4N]2SiF6 single-crystal under green conditions

The present study's purpose is to develop and improve a few reactions that are considered simple, more effective, cleaner, and less costly. In fact, a new (Inorganic-Organic) hybrid single-crystal tetramethylammonium hexafluorosilicate [(CH3)4N]2SiF6, as a heterogeneous ecological catalyst was used, which is found to be effective for the synthesis of tetrahydro-11H-benzo[a]xanthen-11-one via condensation of aromatic aldehydes, β-naphthols with dimedone in EtOH an reflux. The product synthesised were analysed and confirmed by melting point and different spectroscopic techniques FT-IR, 13C NMR, 1H NMR and Elemental Analysis. Furthermore, under ideal conditions of 3.44 mol% [(CH3)4N]2SiF6 and 2 ml EtOH as eco-solvent at reflux, these diverse compounds were produced in outstanding yields in very short durations. The tetramethylammonium hexafluorosilicate shows very high stability and durability for each reuse.

Community Governance Based on Sentiment Analysis: Towards Sustainable Management and Development

The promotion of community governance by digital means is an important research topic in developing smart cities. Currently, community governance is mostly based on reactive response, which lacks timely and proactive technical means for emergency monitoring. The easiest way for residents to contact their properties is to call the property call center, and the call centers of many properties store many speech data. However, text sentiment classification in community scenes still faces challenges such as small corpus size, one-sided sentiment feature extraction, and insufficient sentiment classification accuracy. To address such problems, we propose a novel community speech text sentiment classification algorithm combining two-channel features and attention mechanisms to obtain effective emotional information and provide decision support for the emergency management of public emergencies. Firstly, text vectorization based on word position information is proposed, and a SKEP-based community speech–text enhancement model is constructed to obtain the corresponding corpus. Secondly, a dual-channel emotional text feature extraction method that integrates spatial and temporal sequences is proposed to extract diverse emotional features effectively. Finally, an improved cross-entropy loss function suitable for community speech text is proposed for model training, which can achieve sentiment analysis and obtain all aspects of community conditions. The proposed method is conducive to improving community residents’ sense of happiness, satisfaction, and fulfillment, enhancing the effectiveness and resilience of urban community governance.

Semantic Navigation of PowerPoint-Based Lecture Video for AutoNote Generation

With the increasing popularity of open educational resources in the past few decades, more and more users watch online videos to gain knowledge. However, most educational videos only provide monotonous navigation tools and lack elaborating annotations. This makes the task of locating interesting contents time consuming. To address this limitation, in this article, we propose a slide-based video navigation tool that is able to extract the hierarchical structure and semantic relationship of visual entities in videos, by integrating multichannel information. Features of visual entities are first extracted from the presentation slides by a novel deep learning framework. Then, we propose a clustering approach to extract hierarchical relationships between visual entities (e.g., formulas, texts, or graphs appearing in educational slides). We use this information to associate visual entities with their corresponding audio speech text, by evaluating their semantic relationship. We present two cases where we use the structured data produced by this tool to generate a multilevel table of contents and notes to provide additional navigation materials for learning. The evaluation experiments demonstrate the effectiveness of our proposed solutions for visual entity extraction, hierarchical relationship extraction, as well as corresponding speech text matching. The user study also shows promising improvement in the autogenerated table of contents and notes for facilitating learning.

IMPLEMENTASI APLIKASI QUIZIZZ SEBAGAI MEDIA EVALUASI ONLINE PADA PEMBELAJARAN TEKS PIDATO PERSUASIF KELAS IX SMPN 10 MALANG

The emergence of several problems in learning persuasive speech materials creates its own challenges for teachers, especially when Distance Learning (PJJ) takes place. Distance Learning (PJJ) has been in place since the coronavirus spread in Indonesia, precisely in March 2020. This research was conducted when PJJ took place at SMPN 10 Malang with the research subjects of all 9H grade students numbering 30 students. The purpose of this study is to find out the results of the implementation of the Quizizz application on persuasive speech text learning. The approach used in this study is the quantitative approach. Research data is obtained from the process of conducting tests, spreading questionnaires, conducting observations, and interviews. The collected data is then processed by doing various calculations including the calculation of the validity of the problem item, the reliability of the problem, the different power of the problem, the difficulty level of the problem, and also the percentage calculation of the spread of the questionnaire. The SPSS application was also used in the study to measure the accuracy of the data. The results of this study revealed that (1) positive responses from students while learning using Quizizz, (2) Quizizz applications can increase student learning interest, (3) increase student achievement outcomes. Based on the results of the spread of questionnaires that state that 70% of students agree or are happy when learning to use the Quizizz application.Overall, the results explained that ninth graders felt more excited and experienced increased grades when using the Quizizz application during PJJ.

Pelatihan Penyampaian Kata Sambutan Bagi Pengajian Remaja Nurussalam Aek Badak

The training for delivering the speech text was carried out because self-confidence among PRN youth still needed to be motivated. The PKM method this time offers training implementation that starts with delivering material and practicing with practice. Manuscripts that have been written by each participant are given a section with a slash (/) as a pause marker. This activity is a continuation of the speech writing training which was carried out on the same object. The expected result is that the skills of PRN youth in honing their public speaking skills can be further improved. In this way, PRN youth can take more part in their environment in various social events.

A Recent Survey Paper on Text-To-Speech Systems

All of us are aware of how important knowledge is and it is also true that mostly the data or knowledge is in the form of books or online articles, or various pdfs i.e. in text format. But not all of us are privileged to read. Some are illiterate, some are blind, and some have reading difficulties. Hence, a form of adaptive technology or procedure that reads digital text aloud, which is called text-to-speech (TTS) was developed. It is occasionally referred to as "read-aloud" technology. Words on a computer or other digital device can be converted into audio using TTS. TTS is particularly beneficial for all those people who have reading difficulties or are illiterate. Also, a significant amount of research has been done and is currently being done on text-to-speech technology. Various technologies, methodologies, and algorithms are used in the various proposed approaches and solutions for TTS. This research presents a systematic review of all those methods which have been proposed and implemented by different active researchers in this field.

Symbolic Domination of the Belt and Road Initiative Program on the Speech Text by Chinese President Xi Jinping

Symbolic Domination of the Belt and Road Initiative Program on the Speech Text by Chinese President Xi Jinping

The Role of Artificial Intelligence Technology on English Language Learning: A Literature Review

The word (AI) stands for artificial intelligence, a computer-based simulation of human intelligence meant to act like humans. AI is one of the driving forces behind the 4.0 industrial revolution, making teaching and learning more accessible in schools. This study aims to understand the function of AI in ELT and examine AI technologies in ELT. This is a library research project. The findings indicate that AI provides a positive learning environment for learning English. Depending on the learner's current level of English, career needs, or hobbies, it has much potential to create a customized environment where students can simultaneously use their senses to learn English. AI boosts practical abilities like writing and offers a trustworthy simulation dialogue platform like spoken English. It maximizes the teaching impact of English in ELT while increasing students' practice ability. With the advancement of technology and platforms, learning English has gotten simpler. Artificial intelligence technology provides the chance to enhance English linguistic competence. Students may comprehend English more quickly because many different learning technologies are available. Students get access to a wide variety of ELT apps that are built on AI technology. These technologies include Google Translate, Text to Speech (TTS), EnglishAble, Orai, Elsa, Chatbot, Duolingo, Neo platforms, and many others. Using a method that computers and mobile devices can use, these intelligent machines can mimic intelligence and make decisions as people do.

Addressing the selection bias in voice assistance: training voice assistance model in python with equal data selection

In recent times, voice assistants have become a part of our day-to-day lives, allowing information retrieval by voice synthesis, voice recognition, and natural language processing. These voice assistants can be found in many modern-day devices such as Apple, Amazon, Google, and Samsung. This project is primarily focused on Virtual Assistance in Natural Language Processing. Natural Language Processing is a form of AI that helps machines understand people and create feedback loops. This project will use deep learning to create a Voice Recognizer and use Commonvoice and data collected from the local community for model training using Google Colaboratory. After recognizing a command, the AI assistant will be able to perform the most suitable actions and then give a response. The motivation for this project comes from the race and gender bias that exists in many virtual assistants. The computer industry is primarily dominated by the male gender, and because of this, many of the products produced do not regard women. This bias has an impact on natural language processing. This project will be utilizing various open-source projects to implement machine learning algorithms and train the assistant algorithm to recognize different types of voices, accents, and dialects. Through this project, the goal to use voice data from underrepresented groups to build a voice assistant that can recognize voices regardless of gender, race, or accent. Increasing the representation of women in the computer industry is important for the future of the industry. By representing women in the initial study of voice assistants, it can be shown that females play a vital role in the development of this technology. In line with related work, this project will use first-hand data from the college population and middle-aged adults to train voice assistant to combat gender bias.

The Effect of Deep Learning Methods on Deepfake Audio Detection for Digital Investigation

Voice cloning methods have been used in a range of ways, from customized speech interfaces for marketing to video games. Current voice cloning systems are smart enough to learn speech characteristics from a few samples and produce perceptually unrecognizable speech. These systems pose new protection and privacy risks to voice-driven interfaces. Fake audio has been used for malicious purposes and is difficult to classify what is real and fake during a digital forensic investigation. This paper reviews the issue of deep-fake audio classification and evaluates the current methods of deep-fake audio detection for forensic investigation. Audio file features were extracted and visually presented using MFCC, Mel-spectrum, Chromagram, and spectrogram representations to further study the differences. Harnessing the different deep learning techniques from existing literature were compared using five iterative tests to determine the mean accuracy and the effects thereof. The results showed a Custom Architecture gave better results for the Chromagram, Spectrogram, and Me-Spectrum images and the VGG-16 architecture gave the best results for the MFCC image feature. This paper contributes to further assisting forensic investigators in differentiating between synthetic and real voices.

The Effect of Synthetic Voice Data Augmentation on Spoken Language Identification on Indian Languages

The Effect of Synthetic Voice Data Augmentation on Spoken Language Identification on Indian Languages

MuSE-SVS: Multi-Singer Emotional Singing Voice Synthesizer That Controls Emotional Intensity

MuSE-SVS: Multi-Singer Emotional Singing Voice Synthesizer That Controls Emotional Intensity

Креолизованность как жанрообразующий признак текстов естественной письменной речи (на материале жанра девичьего альбома)

The paper aims at studying the external (technical) aspect of natural written speech text as a characteristic contributing to the materialization of the text as a sign and, at the same time, to its genre identification. The analysis of a girl’s diary genre reveals the heuristic possibilities of creolization as a genre-relevant property of the formal organization of the text, with its texture synthesized by various semiotic systems. Given the relationship between the graphematic means and the codes of other semiotic systems, two types of texts are identified. These are (1) partially creolized texts with autosemantic relations between the verbal and visual components based on the autonomy of structural elements of the text and (2) completely creolized texts with synsemantic relations between the graphematic means and the elements of other semiotic systems. The functions of a girl’s diary as a creolized text are determined: presentational function, contributing to the author’s self-expression; axiological function, aimed at representing the value dominants of a teenage girl; and gaming, determined by the phatic and dialogical components of the text. The combinations of elements of various semiotic systems contributing to the creation of substantial metagraphemics of the text are investigated. Graphematic means are seen to synthesize with various elements of supragraphemics, making it possible to transform the text contents (variation of font, intensity of graphemes); topographemics, associated with the planar variation of the text (paper constructions that change the spatial design of the text); and chromographemics, associated with the color design of the text.

PRESIDENT JOKO WIDODO'S SPEECH TEXT THROUGH APPRAISAL ANALYSIS

This study is a mixed method approach that combines quantitative and qualitative analysis to complement each other. The purpose of the study is to analyze and explain how appraisal system can be described in terms of attitude in the text of Jokowi’s presidential speech. The data of the research is a transcription of President JW’s speech text in 2015. The text document was collected in 2015. In this research, the Appraisal Theory (Martin and White, 2005) is carried out and focus on the analysis of the Attitude element. Further, this prioritizes the discussion of the choice of attitudes that are utilized in the speech. Results showed that the appraisal system in terms of attitude in the text of Jokowi’s presidential speech in 2015 shows that there are 138 attitudes in the speech text. There are 92 positive items and 46 negative items. All kinds of attitudes are spread almost evenly; Affect, Judgment, and Appreciation. This means President JW dares to express his evaluation not only through judgments and appreciations which were disseminated in the text but also expresses his own feelings proved by the affect. The most dominant attitude system portrayed in JW’s speech text in 2015 is propriety [+] and valuation [+].

AN ANALYSIS OF REPRESENTATIVE IN DONALD TRUMP’S IMPEACHMENT CAMPAIGN

This research is an analysis of representative in Donal Trump’s impeachment campaign. The aim of this reasearch is to know how Donal Trump applied representative to persuade the hearers. How a speech of Donal Trump could affect the audience to do something for him to win the second election. In this study, the researcher used qualitative method, and the data were collected from a speech text. The transcript speech text of Donal Trump’s utterances consisting of representative was the instrument of this research. Then, to find and analyze the representative showed in the text speech, the researcher used a theory of John R. Searle (1969) which explains that representative is one of the illocutionary acts that the utterance with the intention to commit the speaker to do something with the truth of expressed proposition. The representatives found in this research were, 1 representative of stating, 3 representatives of informing, 1 representative of predicting, 1 representative of reporting, 1 representative of describing, 1 representative of complaining, 1 representative of claiming, and 2 representatives of asserting. In this study showed that Donal Trump succeed inciting the hearers making a riot to the capitol by believeing in his statements on his impeachment campaign. How his utterances were manipulative and making the audience agreed to the statements he said. The researcher hopes that this analysis would be useful who would like to learn about speech act and illocutionary act, especially about representative illocutionary act.

The role of valence, dominance, and pitch in perceptions of artificial intelligence (AI) conversational agents’ voices

There is growing concern that artificial intelligence conversational agents (e.g., Siri, Alexa) reinforce voice-based social stereotypes. Because little is known about social perceptions of conversational agents’ voices, we investigated (1) the dimensions that underpin perceptions of these synthetic voices and (2) the role that acoustic parameters play in these perceptions. Study 1 (N = 504) found that perceptions of synthetic voices are underpinned by Valence and Dominance components similar to those previously reported for natural human stimuli and that the Dominance component was strongly and negatively related to voice pitch. Study 2 (N = 160) found that experimentally manipulating pitch in synthetic voices directly influenced dominance-related, but not valence-related, perceptions. Collectively, these results suggest that greater consideration of the role that voice pitch plays in dominance-related perceptions when designing conversational agents may be an effective method for controlling stereotypic perceptions of their voices and the downstream consequences of those perceptions.