openSMILE is a software toolkit, for extracting features from audio data for further processing. It’s free for scientific use. It is used in particular for speech and music signals, with applications in eg. detecting speech presence, speaker identity and speaker emotion.
The developer of openSMILE is audEERING GmbH, a German company. Its source code is available, but not with a free license.
The Language Bank has a hands-on tutorial for using it in speech detection.
Product page on audEERING’s site
Get started -page on Github
