A machine learning perspective on the emotional content of Parkinsonian speech

Highlights：

• We introduce a cross-domain transfer-learning model for speech emotion recognition and apply it to a parkinsonian speech corpus.

• We evaluate whether there is a relationship between Parkinson's condition and emotional scores inferred via our speech emotion recognition model.

• Our model infers more negative emotional characteristics in parkinsonian speech than in healthy speech.

• There is a strong correlation between the severity of the speech impairment and of how sad the parkinsonian speech sounds.