Text to Speech

Prosody has been widely used in many speech-related applications including speaker and word recognition, emotion and accent identification, topic and sentence segmentation, and text-to-speech applications. An important application we investigate is that of identifying question sentences in Arabic monologue lectures. Languages other than Arabic have received a lot of attention in this regard. We approach this problem by first segmenting the sentences from the continuous speech using intensity and duration features. Prosodic features are, then, extracted from each sentence. These features are used as input to decision trees to classify each sentence into either question or non question sentence. Our results suggest that questions are cued by more than one type of prosodic features in natural Arabic speech. We used C4.5 decision trees for classification and achieved 75.7% accuracy. Feature specific analysis further reveals that energy and fundamental frequency features are mainly responsible for discriminating between questions and non-question sentences.

The aim of the present study is to evaluate the effect of MultiFunk, a computer program designed to assist reading, on the reading and spelling proficiency of struggling readers. Fifty‐two below‐average readers and spellers, in grades 5, 6 and 7, were randomly assigned as experimental and control groups (N = 26 + 26). In addition, 114 classmates, who read normally, were included to compare changes in basic literacy development during the intervention. A pre‐test, intervention, post‐test, control‐group design was used to evaluate the effects of the software, using texts to suit the pupils’ own choices and interests. The findings indicate that computerized assistive reading has the potential to aid and support the development of basic literacy skills in a broad group of struggling readers and spellers. Issues concerning the MultiFunk text‐to‐speech technology are discussed.

In this paper, an optimized neural network modelling approach with genetic algorithm for text-to-speech system of Arabic language is adopted. The system employs Genetic Algorithm (GA) to perform the connection weights update. An Arabic database which contains subjects, verbs and complements, and a speech synthesizer, whose objective is to make educational oriented verbal Arab sentences, is developed. Based on parallel

Artificial Intelligence has been in great use when it comes to day to day life. Computer science defines AI research as the study of brilliant agents. In almost any direction one turns today, some form of computer-based information processing technology intrudes, whether to the individual knowingly or not. Artificial Intelligence (AI) has already changed our lifestyle. AI device that perceives its environment and takes actions that maximize its chance of successfully achieving its goals. Input to recommendation algorithm can be a database of user and items and output recklessly will be the recommendations. The user is the input into system by voice or text. This paper presents a new approach for smart search. Overall, in world there are many people who use assistant. The paper presents applications of virtual assistant and also describes provocation of applying virtual assistant technology.

In concatenative Text-to-Speech, the size of the speech corpus is closely related to synthetic speech quality. In this paper, we describe our work on a new corpus-based Bell Labs' TTS system. This encompasses large acoustic inventories with a rich set of annotations, models and data structures for representing and managing such inventories, and an optimal unit selection algorithm that accommodates

This paper describes ongoing research that is being conducted at the University of Algarve involving the design and utilization of information and communication technologies for special needed people. We present a virtual keyboard project that interfaces with a text-to-speech synthesizer. The virtual keyboard contains a built-in dictionary and a simple algorithm that allows people with disabilities to write much faster

In the present world, the elderly and illiterate people find it difficult to fill a deposit or withdrawal form in a bank. They often require the help of literate people to fill the form. To solve this problem, we introduce an automatic for filler that would take the customer speech in Malayalam as input and produces the required form. The system is implemented using python modules that involves text-to-speech conversion, language translation, speech-to-text conversion, pdf generation, etc. The system uses googletrans module for translating Malayalam to English and vice-versa. It also uses gTTS module for text-to-speech conversion. The system uses speech-recognition module for converting speech to text format and reportlab module for producing the required form in PDF format. The input speech is converted to text using Speech-recognition module and the generated text is then used for language translation by googletrans module. The questions to be asked is converted from text to speech gTTS module. The system takes both account number and amount from the customer and generate the form with correct entries in PDF format. The generated form is sent to the counter for further processes.

This paper presents a study of Sinhala syllable structure and an algorithm for identifying syllables in Sinhala words. After a thorough study of the Syllable structure and linguistic rules for syllabification of Sinhala words and a survey of the relevant literature, a set of rules was identified and implemented as a simple, easy-to-implement algorithm. The algorithm was tested using 30,000 distinct words obtained from a corpus and compared with the same words manually syllabified. The algorithm performs with 99.95 % accuracy.

This paper presents methodologies involved in text normalization and diphone preparation for Bangla Text to Speech (TTS) synthesis. A Concatenation based TTS system comprises  basically two modules-  one is natural language processing and the other is Digital Signal Processing (DSP). Natural language processing deals with converting text to its pronounceable  form, called Text Normalization and the diphone selection method based on the normalized text is called Grapheme to Phoneme (G2P) conversion.  Text normalization  issues addressed in this paper include tokenization, conjuncts, null modified characters, numerical words, abbreviations  and  acronyms.  Issues related with diphone preparation include diphone categorization, corpus preparation,  diphone  labeling and  diphone selection. Appropriate rules and algorithms are proposed to tackle all the above mentioned issues.  We developed a  speech synthesizer for Bangla using diphone based concatenative approach which is demonstrated to produce much natural sounding synthetic speech.

Telugu is one of the oldest and popular languages of India, spoken by more than 66 million people, especially in South India. Not much work has been reported on the development of optical character recognition (OCR) systems for Telugu text. Therefore, it is an area of current research. Some characters in Telugu are made up of more than one connected symbol. Compound characters are written by associating modifiers with consonants, resulting in a huge number of possible combinations, running into hundreds of thousands. A compound character may contain one or more connected symbols. Therefore, systems developed for documents of other scripts, like Roman, cannot be used directly for the Telugu language. The individual connected portions of a character or a compound character are defined as basic symbols in this paper and treated as a unit of recognition. The algorithms designed exploit special characteristics of Telugu script for processing the document images efficiently. The algorithms have been implemented to create a Telugu OCR system for printed text (TOSP). The output of TOSP is in phonetic English that can be transliterated to generate editable Telugu text. A special feature of TOSP is that it is designed to handle a large variety of sizes and multiple fonts, and still provides raw OCR accuracy of nearly 98%. The phonetic English representation can be also used to develop a Telugu text-to-speech system; work is in progress in this regard.

Due to its clarity and availability, Internet is broadly utilized in practically all the communication applications. In the recent times, number of utilization dependent on web have been created to make the communication as a more reliable and proficient in nature. Out of this various applications, E-mail is the most broadly utilized and dependable approach to speak with one another. The usage of e-mail is quiet easy and clear for ordinary clients yet with regards to the client with visual deformity, the framework is yet exceptionally hard to utilize. Still the current emailing framework is yet not updated for the utilization of visually impaired. This emerges a noteworthy need to redesign the current framework to make it more valuable for the visually impaired. Subsequently, in this study we present an email framework chipping away at the voice controlling rule for the individuals with visual disability to convey a basic and simple admittance to the email framework. This structure will also supportive for the people with different weaknesses alongside the visually impaired peoples.

In the current situation of technology development within the field of augmented reading and voice assistant systems development, many researchers are writing in this regard for the development of new and effective technologies. The voice assistant are software package agents that may interpret human speech & responds via synthesized voices apple siri, microsoft cortana & google assistant are the foremost common voice assistant & embedded in smartphones .Today the technological advancement is increasing day by day. with the assistance of computer system in which we are able to do solely few task however currently artificial intelligence ,machine learning and few a lot of technologies have created computer system so advance that we are able to implement any type of task. In the present paper, The thought of speech technology extremely encompasses 2 technologies: Synthesizer and recognizer. A speech synthesizer takes as input and produces an audio stream as output. A speech recognizer on the opposite hand does opposite. It takes an audio stream as input and thus turns it into text transcription. The voice may be a signal of infinite information. for this reason, we have got created our own virtual personal assistant just for windows using python 3.6 that is ready to right to use on any windows explorer such as windows 7,8,10.we will use python as a programming language as a result of it have a major libraries which is use to implement command.

In this paper, we present a Text to Speech (TTS) synthesis system for Bangla language using the open-source Festival TTS engine. Festival is a complete TTS synthesis system, with components supporting front-end processing of the input text, language modeling, and speech synthesis using its signal processing module. The Bangla TTS system proposed here, creates the voice data for festival, and additionally extends festival using its embedded scheme scripting interface to incorporate Bangla language support. Festival is a ...

"El desarrollo del presente trabajo de Tesis se realizó con el objetivo de contribuir en la línea de investigación interesada en la generación de formas de interacción hombre-máquina por medio del uso de Agentes Conversacionales animados que se desarrolla en la Maestría en Ciencias en Ciencias de la Computación del ITCM, por lo cual se presenta la tecnología Text To Speech a fin de incorporarse a un agente conversacional animado.
Un sintetizador de voz convierte el lenguaje escrito en habla (Text To Speech), el procedimiento de síntesis de voz consiste de dos fases principales, la primera es el análisis del texto y la segunda es la generación de formas de onda de voz, estas etapas producen información fonética y prosódica.
Se presentan las metodologías de síntesis de voz más empleadas actualmente, las cuales se presentan en tres categorías, la primera es la síntesis concatenativa, dentro de la cual se encuentran la síntesis por selección de unidades, la síntesis de dífonos y la síntesis de dominio especifico, otra metodología existente es la síntesis de formantes, también existen otros métodos como la síntesis articulatoria, la síntesis hibrida y la síntesis basada en HMM (Modelos ocultos de Markov).
La fonética y la fonología son disciplinas dentro de la lingüística, ambas se encargan de estudiar los sonidos del lenguaje. La unidad de estudio de la fonología son los fonemas, los cuales son las unidades más pequeñas de sonido del lenguaje.
Los fonemas de un lenguaje pueden ser clasificados en subgrupos basados en sus apariencias visuales, a estos se subgrupos se les llama visemas, se les considera como la equivalencia visual del los fonemas.
Para realizar el análisis de los atributos que intervienen en el proceso de conversión TTS se utilizó el motor de voz Microsoft Speech SDK 5.1 y la librería pyTTS de Python para desarrollar las pruebas. Los atributos de configuración a considerar involucraron el tono, el volumen y la velocidad, así como las voces empleadas para conformar el análisis, estos parámetros arrojaron información sobre los grafemas introducidos como entrada, la información de salida es interpretada como los valores de fonemas y visemas asociados al texto introducido.
Las pruebas fueron diseñadas considerando los fonemas vocálicos y los consonánticos, a fin de observar las variaciones entre los grafemas dependiendo del contexto y del idioma empleado para los análisis."

According to statistics from the World Health Organization (WHO), at least 285 million people are visually impaired or blindness. Blind people generally have to rely on white canes, guide dogs, screen-reading software, magnifiers, and glasses for navigation and surrounding object detection. Therefore, to help blind people, the visual world has to be transformed into the audio world with the potential to inform them about objects. In this paper, we propose a real-time object detection system to help visually impaired people in their daily life. This system consists of a Raspberry Pi in which YOLO (You Only Look Once) deep learning algorithm is employed. We will use YOLOv3 real-time Object Detection algorithm trained on the COCO dataset to identify the object present before the person. Then the label of the object is identified and then converted into audio by using Google Text to Speech (gTTS), which will be the expected output.

Akan is the principal native language in Ghana. While the language has been the object of a number of linguistic studies, researchers in speech technology, in comparison, have not paid as much attention. With the goal of building a complete text-to-speech system of the Akan language, using Microsoft language technology, the present work details a phone set of Asante-Twi, which is one of the most representative dialects of the Akan language. The suggested phone set is transcribed using the two major international phonetic transcription systems, IPA and X-SAMPA, together with the orthographic equivalent of each phone. As an example of phonetic transcription, we translated the text of an Aesopian fable (The North Wind and the Sun) into Asante-Twi and subsequently transcribed it using both versions of our phone set.

Acoustic and articulatory properties of Ukrainian vowels are investigated in this study and a full set of relevant IPA notations are proposed. The notations are shown in the vowel diagram and the table. The results of the earlier acoustic invariant speech analysis based on special software, auditory and spectrum analysis were used and the results are discussed in the context of general and Ukrainian phonetic laws governing language evolution and acoustic properties of non-stressed vowels in relation to their stressed cognates. Such combined approach resulted in a more detailed vowel inventory than proposed heretofore. The findings of this research contribute to better understanding of Ukrainian language and its special features in comparison with other world languages that may have substantial practical use in various phonetic and translation studies, as well as in modern linguistic technologies aimed at artificial intelligence development, machine translation incorporating text-to-speech conversion, automatic speech analysis, recognition and synthesis, and in other areas of applied linguistics.

QR codes, developed by a Japanese company, have been
around for over fifteen years. With the advent of smart and Web
capable mobile devices, we witness a steady growth of interesting
commercial applications using QR codes. The main objective of
our project is to provide people with a technology that can enable
them to obtain information about anything by simply scanning
the QR code. The scanning result in obtaining the text
information of a particular specimen and this information is
further converted into speech using specialized software.

Gesture recognition is the technique
which is used for the interaction between human
and computer or it can be for any automation
implementation purpose. In this paper, we have
made use of the implemented concepts of basic
gesture recognition system and further we propose
a new attachment to this concept i.e. - to convert
these gestures to text and speech. Here we are
comparing hand gestures with the specified
vocabulary of predefined gestures to recognize the
gesture. The paper covers all the necessary idea
and algorithms which help in converting gestures
to text and Speech. The main aim of this paper is to
study and understand the different Algorithms
and tools used in the entire paper, i.e.- it includes
some implemented algorithms for gesture
recognition as a revision and also the use of
proposed algorithms for converting the gestures to
text and speech which are implemented with help
of many API of different programming languages
can be used.

This paper brings together the development of the first Text-to- Speech (TTS) system for Sinhala using the Festival framework and practical applications of it. Construction of a diphone database and implementation of the natural language processing modules are described. The paper also presents the development methodology of direct Sinhala Unicode text input by rewriting letter-to-sound rules in Festival’s context sensitive rule format and the implementation of Sinhala syllabification algorithm. A Modified Rhyme Test (MRT) was conducted to evaluate the intelligibility of the synthesized speech and yielded a score of 71.5% for the TTS system described.

This paper describes about the work done in building a prototype text to speech system for Sanskrit. A basic prototype text-to-speech is built using a simplified Sanskrit phone set, and employing a unit selection technique, where prerecorded sub-word units are concatenated to synthesize a sentence. We also discuss the issues involved in building a full-fledged text-to-speech for Sanskrit.

This research attempts to propose an approach to make decisions under uncertainty for designing a dialogue manager for Amharic spoken dialogue system, Amharic as under-resourced language. A prototype Amharic Spoken Dialogue System was implemented on hotel and restaurant address information domain for experimentation. Data for this research was collected through a method called Wizard of Oz and domain knowledge is prepared using address search websites. How to design a dialogue manager which is robust for languages especially with low performing Automatic Speech Recognition unit is a fundamental question of this research. Previous studies on designing of spoken dialogue system for under resourced languages focused mainly on Automatic Speech Recognizer. We reviewed methods and frameworks on dialogue management. Design of Partially Observable Markov Decision Processes (POMDP) based dialogue manager, which provides a principled framework to plan under uncertainty, yields robustness. Low performing Spoken Dialogue System (SDS) components, especially the Automatic Speech Recognizer (ASR) were considered the causes of uncertainty. Maintaining multiple hypotheses (evidences) improves the correctness of the dialogue manager. We conducted experiments to test correctness score, error rate and robustness. With a maximum of 6 N-best list and 20 partitions the correctness score grow by 14.19%. Increasing the number of n-best list of 6 reduced the error rate by 5.78% and with 6 n-best list. The belief updated below 0.12 seconds with 20 partitions and 6-nbest list. And the dialogue manager was able to complete a task with an average 8.75 turns by 50% Word Error Rate. The finding from the research illustrates that POMDP-based design approach to dialogue management is robust and possible to develop an improving spoken dialogue system for under-resourced languages.

La fonoterapia, que se basa en la mioterapia orofacial, constituiría un procedimiento donde se actúa sobre un músculo o grupo muscular, a través de ejercicios diversos como preparatorios, pasivos, asistidos o por medio de contracciones isotónicas, isométricas, isocinéticas y auxotónicas y/o en la terapia miofuncional orofacial donde se promueven diversos cambios en el tono muscular, en la movilidad y en los patrones funcionales a través del trabajo directo con la función necesitan ser fundamentadas en principios neurofisiológicos que rigen los receptores de las regiones orofacial y cervical.
En el inicio del presente capítulo discutiremos los aspectos generales de la sensibilidad orofacial, integrada en el Sistema somatosensorial; luego se consignarán específicamente las características de los receptores de las regiones orofacial y cervical, además de remarcar la importancia de cada una de las funciones estomatognáticas y sus implicancias en la mioterapia o terapia miofuncional orofacial.

This research focuses on three technologies; ASR, NLP and TTS, and their relationship with each other within digital conversational systems. It looks at different examples of systems that are currently available (2021), detailing and analysing them within each other to determine how they can be improved in the future to continue the fast growth in voice assistants.

In this paper, we describe the design and develop a database of Indonesian diphone synthesis using speech segment of recorded voice to be converted from text to speech and save it as audio file like WAV or MP3. In designing and develop a database of Indonesian diphone there are several steps to follow; First, developed Diphone database includes: create a list of sample of words consisting of diphones organized by
prioritizing looking diphone located in the middle of a word if not at the beginning or end; recording the samples of words by segmentation. ;create diphones made with a tool Diphone Studio 1.3. Second, develop system using Microsoft Visual Delphi 6.0, includes: the conversion system from the input of numbers, acronyms, words, and sentences into representations diphone. There are two kinds of conversion (process) alleged in analyzing the Indonesian text-to-speech system. One is to convert the text to be sounded to phonem and two, to convert the phonem to speech. Method used in this
research is called Diphone Concatenative synthesis, in which recorded sound segments are collected. Every segment consists of a diphone (2 phonems). This synthesizer may
produce voice with high level of naturalness. The Indonesian Text to Speech system can differentiate special phonemes like in ‘Beda’ and ‘Bedak’ but sample of other spesific words is necessary to put into the system. This Indonesia TTS system can handle texts with abbreviation, there is the facility to add such words.