No description available.

abstract: The present study describes audiovisual sentence recognition in normal hearing listeners, bimodal cochlear implant (CI) listeners and bilateral CI listeners. This study explores a new set of sentences (the AzAV sentences) that were created to have equal auditory intelligibility and equal gain from visual information. The aims of Experiment I were to (i) compare the lip reading difficulty of the AzAV sentences to that of other sentence materials, (ii) compare the speech-reading ability of CI listeners to that of normal-hearing listeners and (iii) assess the gain in speech understanding when listeners have both auditory and visual information from easy-to-lip-read and difficult-to-lip read sentences. In addition, the sentence lists were subjected to a multi-level text analysis to determine the factors that make sentences easy or difficult to speech read. The results of Experiment I showed that (i) the AzAV sentences were relatively difficult to lip read, (ii) that CI listeners and normal-hearing listeners did not differ in lip reading ability and (iii) that sentences with low lip-reading intelligibility (10-15 % correct) provide about a 30 percentage point improvement in speech understanding when added to the acoustic stimulus, while sentences with high lip-reading intelligibility (30-60 % correct) provide about a 50 percentage point improvement in the same comparison. The multi-level text analyses showed that the familiarity of phrases in the sentences was the primary driving factor that affects the lip reading difficulty. The aim of Experiment II was to investigate the value, when visual information is present, of bimodal hearing and bilateral cochlear implants. The results of Experiment II showed that when visual information is present, low-frequency acoustic hearing can be of value to speech understanding for patients fit with a single CI. However, when visual information was available no gain was seen from the provision of a second CI, i.e., bilateral CIs. As was the case in Experiment I, visual information provided about a 30 percentage point improvement in speech understanding. / Dissertation/Thesis / Doctoral Dissertation Speech and Hearing Science 2015

Three experiments examined normal-hearing and cochlear-implant listeners' abilities to perceive and use talker-specific information in the speech signal. In Experiment 1 voice similarity judgments were gathered from normal-hearing listeners to maximize variability across talkers used in Experiment 2. These judgments were submitted to a multidimensional scaling (MDS) analysis; this solution was used to select the talkers of Experiment 2. Experiment 2 was an approximate replication of Nygaard and Pisoni's (1998) work. In this study cochlear-implant and normal-hearing listeners were trained to recognize 6 different voices. The cochlear-implant users recognized the voices with 59.31% accuracy and the normal-hearing listeners achieved 92.64% accuracy. After training the listeners completed a sentence recognition task in noise. In the task 6 familiar talkers spoken half of the sentences and 6 novel talkers spoke the other half. It was predicted that sentences spoken by the familiar talkers would be more accurately perceived than those spoken by the novel talkers. However, there was no difference in accuracy, nor was there a difference in performance across the groups of listeners. The factors contributing to these null results were discussed at length. Experiment 3 gathered voice similarity judgments from the normal-hearing and cochlear-implant listeners of Experiment 2. These data were submitted to both classical and weighted MDS analyses. The voice maps showed notable differences in the perceptual spaces of the two groups of listeners. The participant space yielded from the weighted MDS showed great variation across all of the participants' judgments, but no clear trend supporting the listeners' group membership. In conclusion, despite listening via a constrained, electric signal, the cochlear-implant users were trained to recognize voices with notable accuracy (as were the normal-hearing listeners). Nevertheless, Experiment 2 failed to provide insight into talker familiarity's effect on the sentence recognition skills of cochlear-implant and normal-hearing listeners. These results are contrary to research with normal-hearing listeners that suggests talker familiarity facilitates speech processing in noise. The present studies did show, though, that cochlear-implant users appear to perceive and use talker-specific information differently than normal-hearing listeners.

No description available.

No description available.

An important core technology in the development of human language technology applications is an automatic morphological analyser. Such a morphological analyser consists of various modules, one of which is a tokeniser. At present no tokeniser exists for Afrikaans and it has therefore been impossible to develop a morphological analyser for Afrikaans. Thus, in this research project such a tokeniser is being developed, and the project therefore has two objectives: i)to postulate a tag set for integrated tokenisation, and ii) to develop an algorithm for integrated tokenisation. In order to achieve the first object, a tag set for the tagging of sentences, named-entities, words, abbreviations and punctuation is proposed specifically for the annotation of Afrikaans texts. It consists of 51 tags, which can be expanded in future in order to establish a larger, more specific tag set. The postulated tag set can also be simplified according to the level of specificity required by the user. It is subsequently shown that an effective tokeniser cannot be developed using only linguistic, or only statistical methods. This is due to the complexity of the task: rule-based modules should be used for certain processes (for example sentence recognition), while other processes (for example named-entity recognition) can only be executed successfully by means of a machine-learning module. It is argued that a hybrid system (a system where rule-based and statistical components are integrated) would achieve the best results on Afrikaans tokenisation. Various rule-based and statistical techniques, including a TiMBL-based classifier, are then employed to develop such a hybrid tokeniser for Afrikaans. The final tokeniser achieves an ∫-score of 97.25% when the complete set of tags is used. For sentence recognition an ∫-score of 100% is achieved. The tokeniser also recognises 81.39% of named entities. When a simplified tag set (consisting of only 12 tags) is used to annotate named entities, the ∫-score rises to 94.74%. The conclusion of the study is that a hybrid approach is indeed suitable for Afrikaans sentencisation, named-entity recognition and tokenisation. The tokeniser will improve if it is trained with more data, while the expansion of gazetteers as well as the tag set will also lead to a more accurate system / Thesis (M.A. (Applied Language and Literary Studies))--North-West University, Potchefstroom Campus, 2006.

An important core technology in the development of human language technology applications is an automatic morphological analyser. Such a morphological analyser consists of various modules, one of which is a tokeniser. At present no tokeniser exists for Afrikaans and it has therefore been impossible to develop a morphological analyser for Afrikaans. Thus, in this research project such a tokeniser is being developed, and the project therefore has two objectives: i)to postulate a tag set for integrated tokenisation, and ii) to develop an algorithm for integrated tokenisation. In order to achieve the first object, a tag set for the tagging of sentences, named-entities, words, abbreviations and punctuation is proposed specifically for the annotation of Afrikaans texts. It consists of 51 tags, which can be expanded in future in order to establish a larger, more specific tag set. The postulated tag set can also be simplified according to the level of specificity required by the user. It is subsequently shown that an effective tokeniser cannot be developed using only linguistic, or only statistical methods. This is due to the complexity of the task: rule-based modules should be used for certain processes (for example sentence recognition), while other processes (for example named-entity recognition) can only be executed successfully by means of a machine-learning module. It is argued that a hybrid system (a system where rule-based and statistical components are integrated) would achieve the best results on Afrikaans tokenisation. Various rule-based and statistical techniques, including a TiMBL-based classifier, are then employed to develop such a hybrid tokeniser for Afrikaans. The final tokeniser achieves an ∫-score of 97.25% when the complete set of tags is used. For sentence recognition an ∫-score of 100% is achieved. The tokeniser also recognises 81.39% of named entities. When a simplified tag set (consisting of only 12 tags) is used to annotate named entities, the ∫-score rises to 94.74%. The conclusion of the study is that a hybrid approach is indeed suitable for Afrikaans sentencisation, named-entity recognition and tokenisation. The tokeniser will improve if it is trained with more data, while the expansion of gazetteers as well as the tag set will also lead to a more accurate system / Thesis (M.A. (Applied Language and Literary Studies))--North-West University, Potchefstroom Campus, 2006.

Speech audiometry is considered an essential tool in the assessment of hearing, not only to validate pure tone measurements, but also to indicate speech perception as a critical communicative function. The use of sentence material in the assessment of speech perception has great value as it simulates – more closely than single words – the type of speech stimuli that listeners are confronted with on a daily basis. In South Africa, speech recognition (reception and discrimination) abilities are most commonly assessed through the use of single words, presented by monitored live voice, a practice sternly criticized in the literature. Furthermore, speech recognition is commonly evaluated in an ideal (quiet) listening environment. This method gives an incomplete impression of a patient’s auditory performance, since everyday listening situations are often characterised by the presence of background noise that influences comprehension of speech. The present study was therefore launched with the aim to develop a reliable measure of speech recognition in noise using Afrikaans sentence material. The development of the test was conducted in three phases. The first phase entailed the compilation of culturally valid, pre-recorded Afrikaans sentence material. During the second phase the uniformity of the recorded sentence collection was improved by determining the intelligibility of each sentence in the presence of noise and eliminating sentences that were not of equivalent difficulty in this regard. The objective of the third phase was to arrange the sentence material into lists using two different methods of list compilation. The first method involved grouping sentences together based solely on their intelligibility in noise (as assessed in the previous phase). The second method was the well-documented method of compiling phonetically balanced lists. The inter-list reliability of both sets of lists was evaluated in both normal hearing listeners and listeners with a simulated high frequency hearing loss. The results provided valuable information on the process of developing a test of speech recognition in noise, especially in terms of options for list compilation. Findings indicated that lists compiled according to intelligibility in noise showed a higher degree of equivalence than phonetically balanced lists when applied to normally hearing listeners. However, when applied to listeners with a simulated loss, phonetically balanced lists displayed greater equivalence. The developed test provides a means of assessing speech recognition in noise in Afrikaans, and shows potential for application in the assessment of hearing impaired populations, individuals with auditory processing difficulties, and the paediatric population. In addition, the methodology described for the development of the test could provide a valuable guideline for future researchers looking to develop similar tests in other languages. / Dissertation (MCommunication Pathology)--University of Pretoria, 2008. / Speech-Language Pathology and Audiology / MCommunication Pathology / Unrestricted