PhD Dissertations

Permanent URI for this communityhttps://hdl.handle.net/10679/9876

Browse

Browsing by Subject "Adaptation"

Filter results by typing the first few letters
Now showing 1 - 2 of 2
  • Placeholder
    PhD DissertationPublicationMetadata only
    Street vendors' adaptive strategies and the dynamic nature of urban space: Developing an integrative framework in Kano, Nigeria
    Balarabe, Muhammad Kabir; Şahin, Murat; Şahin, Murat; Karahan, Ebru Ergöz; Tekçe, Işılay; Polatoğlu, Ç.; Gülmez, N. Ü.; Department of Design, Technology and Society; Balarabe, Muhammad Kabir
    This thesis attempts to develop a framework that integrates street vendors within the urbanscape. Spaces within the cities are increasingly becoming scares and valuable, resulting in increased competition for them. Coupled with the desire by city administrators and capital interest in create an ideal vision for a city, street vendors are, as a result, pushed away for potentially viable vending locations. However, as is in the global south, vendors are critical services provider to both the formal and informal sector. Additionally, the methods up till now employed to redress these gaps are falling short. Different approached have been taken, such as the design/planning approach, the inclusivity/people/procedural approach, the organisational approach. All have set out strategies for incorporating street vendors, from the mundanity spatial allocation, rethinking vending instrument design, to suggesting urban planning/design process. Despite their attempts, hostile actions and evictions have continued, emphasising their ineffectiveness at integration. Conversely, the number of vendors is increasing despite hostilities. This has created a vicious circle, one counterproductive to the cities, vendors, and users. Within this regard, set to create an alternative, integrative framework informed by the everyday experience of vendors, and the perception of professional and urban users. Taking Kano as a site of investigation, three streets (MM-Way, DEL and SSA) were selected using space syntax methods in conjunction with other qualitative criteria. Street vendors' (ambulatory and non-ambulatory) behavioural mapping was conducted to ascertain the categories and type of vending activities along each street, as well as study their spatial practices and external influences exerted on them. In the philosophical tradition of constructivism, vendors selected through snowball sampling were interviewed and thematically analysed to gain more in-depth knowledge into their adaptive strategies while operating. Users and professional views were sought out to establish the perception of each group towards street vending within the city. The result showed views in a converging, diverging and equivalence range among users and professional. This is translated as an opportunity for co-opting allies. Combining the date with literature regarding integrative approaches, a multidisciplinary framework was developed that placed street vendor's spatial techniques as the entry frame for proposing integration into new and existing geographies. A combination of the result shows that the framework can act as a retrofit, the views espoused by the participant as gauges in evaluating the diversity of participation processes while avoiding tokenism. Finally, street vendors design adaptions were domesticated within the resistance narrative complementary to the conceptions developed within urban sociology.
  • Placeholder
    PhD DissertationPublicationMetadata only
    Using eigenvoices and nearest-neighbours in HMM-based cross-lingual speaker adaptation with limited data
    (2017-08) Sarfjoo, Seyyed Saeed; Demiroğlu, Cenk; Demiroğlu, Cenk; Şensoy, Murat; Uğurdağ, Hasan Fatih; Saraçlar, M.; Güz, Ü.; Department of Computer Science; Sarfjoo, Seyyed Saeed
    Thesis abstract: Cross-lingual speaker adaptation for speech synthesis has many applications, such as use in speech-to-speech translation systems. Here, we focus on cross-lingual adaptation for statistical speech synthesis systems using limited adaptation data. We propose new methods on HMM-based and DNN-based speech synthesis. To that end, for HMM-based speech synthesis we propose two eigenvoice adaptation approaches exploiting a bilingual Turkish-English speech database that we collected. In one approach, eigenvoice weights extracted using Turkish adaptation data and Turkish voice models are transformed into the eigenvoice weights for the English voice models using linear regression. Weighting the samples depending on the distance of reference speakers to target speakers during linear regression was found to improve the performance. Moreover, importance weighting the elements of the eigenvectors during regression further improved the performance. The second approach proposed here is speaker-specific state-mapping which performed signicantly better than the baseline state-mapping algorithm both in objective and subjective tests. Performance of the proposed state mapping algorithm was further improved when it was used with the intra-lingual eigenvoice approach instead of the linear-regression based algorithms used in the baseline system. We propose new unsupervised adaptation method for DNN-based speech synthesis. In this method, using sequence of acoustic features from target speaker, we estimate continuous linguistic features for unlabeled data. Based on objective and subjective experiments, adapted model outperformed the gender-dependent average voice models in terms of quality and similarity.