Kalle Palomäki recently joined ICSI's Speech Group. He is here on ICSI's Finnish visiting program, which is funded by the Finnish Funding Agency for Technology and Innovation through Aalto University and the Helsinki Institute for Information Technology.

Kalle received his doctorate from the Helsinki University of Technology (which later became Aalto University) in 2005. His thesis had two themes. In his initial work, he worked with magnetoencephalography (MEG). MEG measures the magnetic fields caused by electrical currents in the human brain. Kalle and his colleagues measured study participants' neural reactions to simple noises, such as soundbursts and vowels. He was also interested in noise-robust speech recognition and computational auditory scene analysis. An example of this is the cocktail problem: picking a sound or kind of sound - say, a particular speaker - out of background noise. This tends to be easy for human listeners, but not for machines. Generally he's interested in what can be learned from human auditory systems. Recently, he's also become interested in working with general audio, not necessarily just speech. This could include music and city noises. While at ICSI, he will work with Gerald Friedland on speech and non-speech audio in consumer-produced media.