• Overview
  • Research focus lies on integrating linguistic features into statistical models to improve multilingual speech and language processing.
    Link to Google scholar

  • Current projects
    • Sentiment analysis using convolution neural networks
    • Relation classification

  • Previous projects
    • Multilingual bottle-neck features and their applications for rapid language adaptation (funded by Nuance Foundation)
    • Quaero - Research on multimedia and multilingual indexing
    • Code-Switching speech recognition, collaboration with NTU
    • GlobalPhone - Data collection and research on multilingual ASR
    • Rapid Language Adaptation Tools (RLAT)