No description available.

Thesis (M.Sc. (Computer Science)) --University of Limpopo, 2018 / The conversion of speech to text is essential for communication between speech and visually impaired people. The focus of this study was to develop and evaluate an ASR baseline system designed for normal speech to correct speech disorders. Normal and disordered speech data were sourced from Lwazi project and UCLASS, respectively. The normal speech data was used to train the ASR system. Disordered speech was used to evaluate performance of the system. Features were extracted using the Mel-frequency cepstral coefficients (MFCCs) method in the processing stage. The cepstral mean combined variance normalization (CMVN) was applied to normalise the features. A third-order language model was trained using the SRI Language Modelling (SRILM) toolkit. A recognition accuracy of 65.58% was obtained. The refinement approach is then applied in the recognised utterance to remove the repetitions from stuttered speech. The approach showed that 86% of repeated words in stutter can be removed to yield an improved hypothesized text output. Further refinement of the post-processing module ASR is likely to achieve a near 100% correction of stuttering speech Keywords: Automatic speech recognition (ASR), speech disorder, stuttering

IBM, a well-known computer giant, and Nuance, a renowned speech technology firm, have been offering numerous speech recognition applications in the recent years. The connections between these two companies and the automobile, communication, and other eight dominating industries, including banking, electronics, energy/utilities, medical/life science, insurance, media/entertainment, retail travel and transportation, are vastly expanded and flourished. Maturity of these speech technologies drives our lifestyle to a cozy level that we cannot imagine before. In April, 2011, the world class manufacturer Foxconn decided to invest 12 billion US dollars to build iPhone/iPad factories in Brazil, the largest Portuguese speaking country in the world. It is our objective to build a language system that can help us to learn Portuguese, to savor the beauty of their culture, and to widen our vision of travel and living. This thesis investigates the design and implementation strategies for a Portuguese speech recognition system. It utilizes the speech features of the 303 common Portuguese mono-syllables as the major training and recognition methodology. A training database of 10 utterances per mono-syllable is established by applying Portuguese pronunciation rules. These 10 utterances are collected through reading 5 rounds of the same mono-syllables twice with different tones. The first pronounced pattern has high pitch of tone 1, while the second one has falling pitch of tone 4. Mel-frequency cepstral coefficients, linear predicted cepstral coefficients, and hidden Markov model are used as the two syllable feature models and the recognition model respectively. Under the AMD 2.2 GHz Athlon XP 2800+ personal computer and Ubuntu 9.04 operating system environment, correct phrase recognition rates of 87.26% can be reached using phonotactical rules for a 3,900 vocabulary Portuguese phrase database. The average computation time for the Portuguese phrase system is less than 1.5 seconds, and the training time for the systems is about two hours.

Language plays an important role for understanding people, their history, culture and even technology. Many countries of the world have developed the technology of the outer space recently, and Russian is the top of the world. In 1998 Russia further launched Zarya, the first International Space Station (ISS) Module, to the outer space, and was deeply involved in the development of the ISS with the U.S.. Since the end of the World War Two, Russia has been one of the five Permanent Members in the United Nations. And then, she became one of the G8 members, an economical forum of eight industrially advanced nations. Because these informations, it is our objective to build a language system that can help us to learn Russian, to taste the beauty of her culture, and to widen our vision of technologies. This thesis investigates the design and implementation strategies for a Russian speech recognition system. It utilizes the speech features of the 514 common Russian mono-syllables as the major training and recognition methodology. The mono-syllable is established by applying Russian pronunciation rules. These 12 utterances are collected through reading 6 rounds of the same mono-syllables twice with different tones. The first pronounced pattern has high pitch of tone 1, while the second one has falling pitch of tone 4. Mel-frequency cepstral coefficients, linear predicted cepstral coefficients, and hidden Markov model are used as the two syllable feature models and the recognition model respectively. Under the AMD 2.2 GHz Athlon XP 2800+ personal computer and Ubuntu 9.04 operating system environment, correct phrase recognition rates of 86.90% and 94.83% can be reached using phonotactical rules for a 3,900 vocabulary Russian phrase database for TORFL (Test of Russian as a Foreign Language) and a 600 person name database for Russian. The average computation time for each system is less than 1.5 seconds, and the training time for the systems is about three hours.

Arab world is one of the most spectacular regions in the earth, especially for her over 2,800 year history, Islamic religion and magnificent culture. She consists of 24 countries and territories where people speak Arabic. The population of Arabic speaking people is approximately 221 million, and ranked the fourth according to the 2009 statistics by Summer Institute of Linguistics, USA. Since 1973, petroleum embargoes, imposed by the Arab world, have influenced global economy and hurt national security seriously. This kind of fossil energy is still irreplaceable until efficient green energy alternative becomes feasible. It is our objective to build a language system that can help us to learn Arabic, to appreciate the beauty of her culture, and to widen our vision of religions. This thesis investigates the design and implementation strategies for an Arabic speech recognition system. It utilizes the speech features of the 302 common Arabic mono-syllables as the major training and recognition methodology. A training database of 10 utterances per mono-syllable is established by applying Arabic pronunciation rules. These 10 utterances are collected through reading 5 rounds of the same mono-syllables twice with different tones. The first pronounced pattern has high pitch of tone 1, while the second one has falling pitch of tone 4. Mel-frequency cepstral coefficients, linear predicted cepstral coefficients, and hidden Markov model are used as the two syllable feature models and the recognition model respectively. Under the AMD 2.2 GHz Athlon XP 2800+ personal computer and Ubuntu 9.04 operating system environment, correct phrase recognition rates of 86.31% and 93.90% can be reached respectively using phonotactical rules for a 3,600 vocabulary Arabic phrase database and a 590 person name database for Arabic figures. The average computation time for each system is less than 1 second, and the training time for the systems is about two hours.

The European Union (EU) established on November 1, 1993, according to the Maastricht Treaty signed on February 7, 1992. This economic and political community consists of 27 member states, primarily located in Europe. She operates through a supranational and intergovernmental system, including the European Commission, the Council, the Parliament and the Central Bank, to transfer herself from the joint economic development regions to the single market of economic and political integration. Italy is one of the six founding countries of the EU, also one of the G8 members, the eight industrially advanced nations in the world, and playing a force to be reckoned with. It is our objective to build a language system that can help us to learn Italian more effectively, to promote our competency of intercultural understanding, and to widen our vision of travel and living. This thesis investigates the design and implementation strategies for an Italian speech recognition system. It utilizes the speech features of the 370 common Italian mono-syllables as the major training and recognition methodology. A training database of 10 utterances per mono-syllable is established by applying Italian pronunciation rules. These 10 utterances are collected through reading 5 rounds of the same mono-syllables twice with different tones. The first pronounced pattern has high pitch of tone 1, while the second one has falling pitch of tone 4. Mel-frequency cepstral coefficients, linear predicted cepstral coefficients, and hidden Markov model are used as the two syllable feature models and the recognition model respectively. Under the AMD 2.2 GHz Athlon XP 2800+ personal computer and Ubuntu 9.04 operating system environment, correct phrase recognition rates of 88.35% and 89.32% can be reached using phonotactical rules for a 4,000 vocabulary Italian phrase database and a 3,304 word database for Italian Language Proficiency Test. The average computation time for each system is less than 1.5 seconds, and the training time for the systems is about two hours.

The Republic of Turkey, founded in 1923, is a well-known ancient country with abundant cultural heritage and great junction location of the Asian and European Continents. Istanbul is the largest city of this country with her old name Constantinople or Byzantium. She was established by Constantinus I Magnus in A.D. 330 during the era of the Roman Empire, to serve as a well-fortified castle like Rome. Numerous attractions on historical architecture, ancient music, gourmet cuisine, and art collections can be explored and appreciated. It is our objective to build a language system that can help us to learn Turkish, to savor the beauty of her culture, and to widen our vision of travel and living. This thesis investigates the design and implementation strategies for a Turkish speech recognition system. It utilizes the speech features of the 395 common Turkish mono-syllables as the major training and recognition methodology. A training database of 12 utterances per mono-syllable is established by applying Turkish pronunciation rules. These 12 utterances are collected through reading 6 rounds of the same mono-syllables twice with different tones. The first pronounced pattern has high pitch of tone 1, while the second one has falling pitch of tone 4. Mel-frequency cepstral coefficients, linear predicted cepstral coefficients, and hidden Markov model are used as the two syllable feature models and the recognition model respectively. Under the AMD 2.8 GHz Athlon X2 2400 personal computer and Ubuntu 9.04 operating system environment, correct phrase recognition rates of 87.29% can be reached using phonotactical rules for a 3,644 vocabulary Turkish phrase database. The average computation time for the each system is less than 1.5 seconds, and the training time for the systems is about two hours.

Britain established the status of maritime hegemony in 1588. The English language along with the British colonized activities was spread to North America, India, Africa and Australia. After the end of World War I in 1918, the U.S. became the most powerful nation in the world economy. And at the same time, the world financial center was shifted to New York from London. In 1945, the World War II ended, the U.S. further played indispensable role in each aspect of international politics, economy and technologies. The United Nation, founded on October 24, 1945, adopted English, Chinese, French, Spanish, Arabic as well as Russian as the six working languages. These historical events facilitated a succession of language expansion and caused English to be the most widely used international language. Beside the political, economic and technological superiority, Britain owns the largest comprehensive museum in the globe, the British Museum. This Museum was located in London, built in 1753, and more than 13 million cultural relics of archaeology from around the world were collected. Her cultural resources are remarkably rich. It is our objective to build a language system that can help us to learn English more effectively and to widen our vision of living at the same time. This thesis investigates the recognition rate improvement strategies for an English speech recognition system. It utilizes the speech features of the 989 common English mono-syllables as the major training and recognition methodology. A training database is established by reading each mono-syllable 14 rounds. Each one of the 989 mono-syllables is consecutively read with two different tones at alternate rounds. The odd pronounced rounds have high pitch of tone 1, while the even rounds have falling pitch of tone 4. The pitch period frame method is applied for enhancing the accuracy of end point detection. Mel-frequency cepstral coefficients, linear predictive cepstral coefficients, and hidden Markov model are used as the two feature models and the recognition model respectively. The number of HMM states is adjusted to 10 and the phonotactical rule is used for the recognition rate improvement. Under the Core ™ i5 CPU M450 notebook computer with 2.4GHz clock rate and Fedora 14 operating system environment, a 92.94% correct phrase recognition rate can be reached for a 6,812 English phrase database. The average computation time for each phrase is within 1.5 seconds.

In this thesis, both Turkish and Tamil, a language spoken in southern India and Sri Lanka, are studied in addition to Mandarin Chinese. It is hoped that the history, culture, and economy behind each language can be acquainted, tasted and appreciated during the learning process. In the ancient Chinese Han and Tang Dynasties, the ¡§Silk Road¡¨ played the most magnificent role to connect among the Oriental China, the Western Turkey and the Southern India as the international trading corridor. In this modern era, Turkey and India are both the most important cotton exporting countries. Moreover, China, Turkey and India have been showing their potential to the newly emerging markets in the world. Therefore, a trilingual speech recognition system is developed and implemented to help us to learn Chinese, Turkish and Tamil, as well as to enhance our understanding to their history and culture. In this trilingual system, linear predicted cepstral coefficients, Mel-frequency cepstral coefficients, hidden Markov model and phonotactics are used as the two syllable feature models and the recognition model respectively. For the Chinese system, a 2,699 two-syllable words database is used as the training corpus. For the Turkish and Tamil systems, a database of 10 utterances per mono-syllable is established by applying their pronunciation rules. These 10 utterances are collected through reading 5 rounds of the same mono-syllables twice with tone 1 and tone 4. The correct rates of 88.30%, 84.21%, and 88.74% can be reached for the 82,000 Chinese, 30,795 Turkish, and 3,500 Tamil phrase databases respectively. The computation time for each system is within 1.5 seconds. Furthermore, a trilingual language-speech recognition system for 300 common words, composed of 100 words from each language, is developed. A 98% correct language-phrase recognition rate can be reached with the computation time less than 2 seconds.

In this thesis, an audio-video database for speaker recognition is constructed using a digital camcorder. Motion pictures of fifteen hundred speakers are recorded in three different sessions in the database. For each speaker, 20 still images per session are also derived from the video data. It is hoped that this database can provide an appropriate training and testing mechanism for person identification using both voice and face features.