Libraries & frameworks
opensmile
audeering/opensmile
openSMILE is a complete open-source toolkit for extracting audio features from speech and music, used in recognition, emotion analysis and related research.
Topics
Related projects
basic-pitch6kA lightweight Python library and CLI from Spotify that transcribes audio to MIDI using a small neural network, with polyphonic note and pitch-bend detection.
audio3kTorchAudio, PyTorch's official library for loading audio datasets and applying transforms such as spectrograms, MFCC, and resampling in machine learning pipelines.
audiomentations2.3kAudiomentations is a Python library for audio data augmentation, offering composable transforms that help audio machine-learning models generalize beyond the lab.
spotlight1.3kSpotlight is a Python library that turns dataframes into interactive browser views for exploring images, audio, text and other unstructured data, helping ML teams find clusters and data issues.
muspy529MusPy is a Python toolkit for symbolic music generation, offering dataset management, MIDI/MusicXML I/O, common music representations and objective evaluation tools for music generation systems.
neural-amp-modeler-lv2551An LV2 plugin that runs neural amp models through the NeuralAudio engine, bringing machine-learning guitar amp and pedal tones into LV2 hosts.