search

audeering / opensmile

The Munich Open-Source Large-Scale Multimedia Feature Extractor
https://audeering.github.io/opensmile/
Other
553 stars 74 forks source link

Extract emotions from speech #45

Closed dstefa02 closed 2 years ago

dstefa02 commented 2 years ago

Hello,

Is it possible to extract emotions from speech using opensmile (e.g. happy, anger, etc)? In the main page, it states that this toolkit can be also used for emotion recognition, but it is not clear whether this functionality is available (e.g. pretrained model) and how this can be achieved.

Any help or advise would be greatly helpful. Thanks

chausner-audeering commented 2 years ago

openSMILE is purely an audio feature extractor and does not come with any pre-trained models. In order to use openSMILE for emotion recognition, a model needs to be trained first on the output of openSMILE and some training set.

We (audEERING) do offer devAIce as a commercial solution for emotion recognition and other audio intelligence tasks. These products come with pre-trained models.