ptnghia-j/ChordMiniApp
Music Analysis, Chord Recognition, Beat Tracking, Guitar Diagrams, Piano Visualizer, Lyrics Transcription Application, context-aware LLM inference for analysis from uploaded audio and YouTube video
What it solves
ChordMini is an open-source music analysis tool designed to automate the process of extracting musical structure from audio. It solves the difficulty of manually transcribing chords, beats, and melodies from songs, providing musicians and learners with visual aids like guitar diagrams and piano rolls.
How it works
The application uses a combination of machine learning models and external APIs to analyze audio files or YouTube videos. It employs a Python backend to run beat detection (via Madmom and Beat-Transformer) and chord recognition (via Chord-CNN-LSTM and BTC variants). For advanced features, it integrates the Google Gemini API for Roman numeral analysis and lyrics translation, and uses SongFormer for song segmentation. Experimental melody transcription is handled by a separate Sheet Sage service.
Who it’s for
It is designed for musicians, music students, and producers who need quick chord progressions, beat tracking, and visual guides for playing along with a song.
Highlights
- Comprehensive Analysis: Performs chord recognition, beat tracking, and song segmentation (intro, verse, chorus, etc.).
- Interactive Visualizers: Includes a real-time piano roll with MIDI export and interactive guitar chord diagrams.
- AI-Assisted Lead Sheets: Combines synchronized lyrics with a Gemini-powered chatbot for contextual music analysis.
- Multi-Source Input: Supports direct audio uploads and YouTube URL integration.
- Experimental Melody Transcription: Can estimate melodic lines and render them as sheet music.