The following background is required to join our lab.
Fundamental training in mathematics and signal processing is essential.
Higher-level knowledge in stochastic processes, time series analysis,
probability theory, mathematical statistics, numerical analysis,
differential geometry, information theory, pattern information theory,
machine learning, artificial intelligence, natural language
processing, phonetics/phonology, linguistics, music composition,
speech coding, speech recognition, speech synthesis, spoken dialog,
computer graphics, virtual reality, and anthropomorphic agent is
Particularly, knowledge and experiences in Hidden Markov Models,
Dynamic Programming, Linear Predictive Coding, Spectral Analysis,
Neural Networks, Speech Recognition Algorithms, and Statistical
Estimation (EM-algorithm) will be helpful.
For music information processing, basic knowledge in music theory is
necessary. Well-organized knowledge in harmony theory, counterpoint,
history of classical music, and music composition is helpful.
Keyboard playing and other music instrument playing skills, and
MIDI-based desk-top music skill are useful.
Some computer skill is essential.
Fundamental computer skills for Linux operations, C programming,
and shell script programming are essential.
C++ programming and practical script languages such as perl, awk,
ruby, Tcl/Tk, html, etc. are also suggested.
Matlab may be helpful in limited purposes but is not sufficient in speed
and flexibility for intensive research at our lab.
X programming and OpenGL programming skills are also appreciated in
the area of anthropomorphic spoken dialogue agent (digital human).
Document and slide preparation with TeX in combination with
tgif and gnuplot is recommended.