TART: A Modular Tool for Technique-Aware Audio-to-Tablature Guitar Transcription
TART claims stronger guitar tab transcription by adding technique labels and audio-aware fingering to the pipeline.
The paper describes a four-stage system that turns guitar audio into MIDI, detects expressive techniques, assigns string-fret positions, and generates tablature. It targets slides, bends, percussive hits, incorrect fingering, and noisy recordings as the main failure points for prior systems. In zero-shot tests across GuitarSet, EGDB, and noisy variants of both, TART reports 81.35% audio-to-MIDI F50 and 71.8% string-fret Tab F1, beating the cited best baselines by 6.67 and 8.5 points. The authors say it is the first framework to generate guitar tablature with both fingering and expressive technique annotations directly from guitar audio. ArXiv · AI/CL/LG's note
The paper describes a four-stage system that turns guitar audio into MIDI, detects expressive techniques, assigns string-fret positions, and generates tablature. It targets slides, bends, percussive hits, incorrect fingering, and noisy recordings as the main failure points for prior systems. In zero-shot tests across GuitarSet, EGDB, and noisy variants of both, TART reports 81.35% audio-to-MIDI F50 and 71.8% string-fret Tab F1, beating the cited best baselines by 6.67 and 8.5 points. The authors say it is the first framework to generate guitar tablature with both fingering and expressive technique annotations directly from guitar audio. ArXiv · AI/CL/LG's note
score 4