Colloquium Joep Boers

Humans are very well able to detect aggression in verbal expressions. Knowledge of
the relation between emotions and acoustic features in speech may be of use
improving, e.g., speech recognition. Currently, Sound Intelligence is developing
the next generation of aggression detectors. Those systems are aimed at not only
detecting, but also classifying aggression.

In order to come to aggression classification I focused on speech production.
Parallel to human speech production, I implemented a vocoder, which I used to set up
a database of synthesized vowels. In contrast to former research (normally
conducted on genuine, rich and labeled data), I defined cues and subsequently
synthesized vowels intended to exhibit gradations of aggression. By means of a
psycho-acoustic experiment this approach, and thus the vocoder, has proven to be
scientifically justified. Still, aggression classification needs much more further
research - and it is my belief that nonlinear analysis might be very useful here
(literature shows very interesting progress) - but either way, a vocoder, like the
one used in this work, is believed to be complementary to current research
approaches.