Tuomas Virtanen research homepage

I am professor at Tampere University, where I am leading the Audio Research Group.

Research interests

  • Audio signal processing:
    • content analysis of audio, machine listening
    • sound source separation, object-based audio coding
    • non-negative spectrogram decompositions, sinusoidal modeling
  • Computational analysis of everyday soundscapes
    • Sound event detection, acoustic scene classification
    • Active learning, zero-shot learning
  • Speech processing:
    • noise-robust automatic speech recognition, speech separation,
    • speech enhancement, speaker recognition, voice activity detection
  • Music signal processing:
    • music content analysis
    • source separation of music signals, automatic transcription of music
    • musical instrument recognition, singing analysis
  • Acoustic pattern recognition: speech, music and environmental sounds
  • Machine learning, statistical modeling, pattern recognition, mathematical optimization


Audio demonstrations

Audio processing software

DCASE: Challenge on Detection and Classification of Acoustic Scenes and Events



Contact information:

Tuomas Virtanen
Tampere University
P.O. Box 553
FI-33101 Tampere

Tel. +358 401981308