Skip to main content
eScholarship
Open Access Publications from the University of California

Scholarly Works (2 results)

  • Thesis
  • Peer Reviewed
UC Berkeley Electronic Theses and Dissertations (2010)

In this work, I investigated structured approaches to data selection for speaker recognition, with an emphasis on information theoretic approaches, as well as approaches based on speaker-specific differences that arise from speech production. These approaches rely on the investigation of speaker discriminability measures that detect speech regions that result in high speaker differentiation. I also attempted to understand why certain data regions result in better speaker recognition system performance.

The knowledge gained from the speaker discriminability measures was used to implement an effective data selection procedure, that allows for the prediction of how well a speaker recognition system will behave without actually implementing the system. The use of speaker discriminability measures also leads to data reduction in speaker recognition training and testing, allowing for faster modeling and easier data storage, given that the latest speaker recognition corpora uses hundreds of gigabytes.

In particular, I focused primarily on Gaussian Mixture Model- (GMM) based speaker recognition systems, which comprise the majority of current state-of-the-art speaker recognition systems. Methods were investigated to make the speaker discriminability measures easily obtainable, such that the amount of computational resources required to extract these measures from the data would be significantly less in comparison to the computational resources required to run entire speaker recognition systems to determine what regions of speech are speaker discriminative.

Upon selecting the speech data using these measures, I created new speech units based on the data selected. The speaker recognition performances of the new speech units were compared to the existing units (mainly mono-phones and words) standalone and in combination. I found that in general, the new speech units are more speaker discriminative than the existing ones. Speaker recognition systems that use the new speech units as data in general outperformed systems using the existing speech units. This work, therefore, outlines an effective approach that is easy to implement for selecting speaker discriminative regions of data for speaker recognition.

    Cover page: Structured Approaches to Data Selection for Speaker Recognition
    • Article
    • Peer Reviewed
    UC Irvine Previously Published Works (2020)
    The COVID-19 pandemic has required greater minute-to-minute urgency of patient treatment in Intensive Care Units (ICUs), rendering the use of Randomized Controlled Trials (RCTs) too slow to be effective for treatment discovery. There is a need for agility in clinical research, and the use of data science to develop predictive models for patient treatment is a potential solution. However, rapidly developing predictive models in healthcare is challenging given the complexity of healthcare problems and the lack of regular interaction between data scientists and physicians. Data scientists can spend significant time working in isolation to build predictive models that may not be useful in clinical environments. We propose the use of an agile data science framework based on the Scrumban framework used in software development. Scrumban is an iterative framework, where in each iteration larger problems are broken down into simple do-able tasks for data scientists and physicians. The two sides collaborate closely in formulating clinical questions and developing and deploying predictive models into clinical settings. Physicians can provide feedback or new hypotheses given the performance of the model, and refinement of the model or clinical questions can take place in the next iteration. The rapid development of predictive models can now be achieved with increasing numbers of publicly available healthcare datasets and easily accessible cloud-based data science tools. What is truly needed are data scientist and physician partnerships ensuring close collaboration between the two sides in using these tools to develop clinically useful predictive models to meet the demands of the COVID-19 healthcare landscape.
      Cover page: Agile clinical research: A data science approach to scrumban in clinical medicine
      Top