Browsing by Author "Bredmar, Fredrik"

Filter results by typing the first few letters
Now showing 1 - 2 of 2
  • No Thumbnail Available
    Item
    Scalable Machine Learning for Big Data
    (2014-09-22) Bredmar, Fredrik; Andersson, Emanuel; Bogren, Emil; Göteborgs universitet/Institutionen för data- och informationsteknik; University of Gothenburg/Department of Computer Science and Engineering
    We describe each step along the way to create a scalable machine learning system suitable to process large quantities of data. The techniques described in the report will aid in creating value from a dataset in a scalable fashion while still being accessible to non-specialized computer scientists and computer enthusiasts. Common challenges in the task will be explored and discussed with varying depth. A few areas in machine learning will get particular focus and will be demonstrated with a supplied case-study using weather data courtesy of the Swedish Meteorological and Hydrological Institute.
  • No Thumbnail Available
    Item
    Speech-to-speech translation using deep learning
    (2017-03-17) Bredmar, Fredrik; Göteborgs universitet/Institutionen för data- och informationsteknik; University of Gothenburg/Department of Computer Science and Engineering
    Current state-of-the-art translation systems for speech-to-speech rely heavily on a text representation for the translation. By transcoding speech to text we lose important information about the characteristics of the voice such as the emotion, pitch and accent. This thesis examine the possibility of using an LSTM neural network model to translate speech-to-speech without the need of a text representation. That is by translating using the raw audio data directly in order to persevere the characteristics of the voice that otherwise get lost in the text transcoding part of the translation process. As part of this research we create a data set of phrases suitable for speech-to-speech translation tasks. The thesis result in a proof of concept system which need to scale the underlying deep neural network in order to work better.