Here are the pretty good comparison and evaluation of various audio feature extraction toolboxes.
Main conclusion based on the paper mention in the link below:
1) Essentia: Full function workflow environment for high and low-level features, facilitating audio input, preprocessing and statistical
analysis of output. Written in C++, with Python binding and
export data in YAML or JSON format.
2) Marsyas: Full real-time audio processing standalone framework
for data flow audio processing with GUI and CLI. This program
includes a low-level feature extraction tool built in C++, with an ability
to perform machine learning and synthesis within the framework.The
feature extraction aspects have also been translated to
Vamp plugin format
3)YAAFE Low level feature extraction library designed for computational
efficiency and batch processing by utilizing data flow
graphs, written in C++ with a CLI and bindings for Python and
Matlab
You can find the more details about Graph of Percentage Coverage of Multiple Feature Sets and Graph of Computational Time of Feature Extraction
Tools in the
link.