Dr. Nasser Mohammadiha
- Statistical signal processing with applications to speech processing and image processing
- Machine learning for signal processing
- Some of the topics that I am interested to work on include: model-based speech enhancement (supervised and unsupervised), blind source separation, nonnegative matrix factorization (NMF) and dynamic NMF, hidden Markov models (HMM) and Kalman filtering, Bayesian inference, latent variable analysis and nonparametric learning, and deep learning.
I received my M.Sc. degree in Electronics Engineering from Sharif University of Technology, Tehran, IRAN. I completed my PhD studies at the Communication Theory Lab at KTH Royal Institute of Technology, Stockholm, Sweden in Oct. 2013 (PhD Thesis) under the supervision of Prof. Arne Leijon and Prof. Bastiaan Kleijn. Since Nov. 2013, I am a postdoctoral researcher at the Dept. of Medical Physics and Acoustics at the University of Oldenburg, Germany.
From 2009 until 2012, I was with the European project AUDIS (Digital Signal Processing in Audiology) during which I visited Ruhr-Universität Bochum, Germany, and University of Southampton, UK for short time. I visited Prof. Paris Smaragdis's group at University of Illinois at Urbana-Champaign, USA from October 2012 until June 2013. I have taught some undergrad courses at Azad University, Iran for two years, which covered signal theory, microprocessors, introduction to devices and circuits, introduction to communications and signal processing. I have also been actively working on designing DSP boards using FPGA, microprocessors and associated softwares from 2004 to 2008.
- I got the Best Student Paper award of European Signal Process. Conf. (EUSIPCO) 2013 for the contribution: N. Mohammadiha, W. B. Kleijn, A. Leijon, "Gamma Hidden Markov Model as a Probabilistic Nonnegative Matrix Factorization ," in European Signal Process. Conf. (EUSIPCO), 2013.