A Lightweight Instrument-Agnostic Model for Polyphonic Note Transcription¶
Summary¶
Spotify's lightweight neural network for converting audio to MIDI. Uses a harmonic CQT representation fed through a small CNN to predict multi-pitch activations and note events with pitch bends. Designed for consumer use: ~15MB model, runs on CPU in near real-time, works on any pitched instrument (voice, violin, guitar, etc.). Instrument-agnostic by design — no instrument-specific training needed.
Key Claims¶
- A small harmonic CQT + CNN architecture can achieve competitive polyphonic transcription
- Instrument-agnostic training works: the model generalizes across instrument timbres
- Pitch bend detection is critical for realistic transcription of non-keyboard instruments
- 15MB model size enables on-device consumer deployment
Related¶
- ../entities/basic-pitch — the tool (pip installable)
- ../entities/crepe — monophonic pitch estimation, complementary
- ../entities/mt3 — multi-instrument transcription, different approach
- ../concepts/musicxml-tab-notation — output can feed into tab notation