Train no-reference speech quality estimators with multiple datasets via learned, per-dataset alignments.

Manipulating of speech files (pitch, spectral) with Praat. Listening test and data analysis of subjective perceptions of speaker characteristics.

Java GUI for labeling speech files in terms of social speaker characteristics (likable, attractive, competent, childish, etc.).

Automatic classification of audio files based on the perceived level of reverberation

ExperimentMFC port for web