emotionSpeechRecognition

Emotion Speech Recognition is a problem inside affective computing that teach machines how is someone emotionally behaving thanks to speech signals. Speech signals are important biomarkers due to its significant combination of neuromuscular, physiological and acoustic features.

This project pretends to analyse and compare 2 different datasets using a model based on DepAnxSpeechNet. Data is available in their original sources TESS and BerlinEMODB. The results are promising and increasing the level of previous studies.

Many thoughts come from the difference analysing German and English, even for machines is easier to get the mood of a speaker in english than in german.