My research was funded by the Deutsche Forschungsgemeinschaft by a project about phonetic reductions and elaborations in emotional speech. The main aim of the project was the creation of an emotional database for German and analysis of acoustical correlates of emotional speech.
I was specially involved in the verification of the perceptive relevance of typical features of the emotions by means of speech synthesis.
In order to generate stimuli for my perception-tests I've written a program to simulate emotional speech with formant synthesis - I call it emoSyn
Since working for Deutsche Telekom (until 2018) and now at audEERING my fields of interest are mainly
- Human-Machine dialog platform architectures
- Ontology based semantic modeling
- Emotional Human Machine interaction
- Automatic speaker characteristics classification
other stuff
-
my master's thesis: simulation of
happiness with speech synthesis (in German)
-
a short summary of the results from my
master's thesis in English (as a zipped
postscript, 424 kB)
-
TTS-tool
emofilt (a tool to modify mbrola-input)
-
I maintain a collection of synthesized emotional speech.
-
Another collection of German speech synthesizer.
- Emotional Applications a taxonomy
-
A collection of demos from TTS-Systems for German.
-
Text2Speech
System Siphon, in German (a complete German tts-system in tcl/tk I've written
a long time ago
mail: Felix Burkhardt