https://github.com/bagustris/sner - v1.0

Added: - able to receive any file format supported by audofile - model was built on scikit-learn MLPregressor - acoustic feature extracted from ComParE 2016 high-level statistical functions (6k+) - acoustic feature is scaled using StandardScaler() inside model (fit during training) - supported arguments: input file, split, duration - output (in order): valence, arousal, dominance, naturalness, in range [-1, 1]

Full Changelog: https://github.com/bagustris/sner/commits/v1.0

- Python
Published by bagustris over 4 years ago