Problem
Research questions and friction points this paper is trying to address.
Automatic Pitch Correction
Harmonic Modeling
Vocal-Accompaniment Context
MIDIBack: Harmony-Aware Singing Pitch Correction via Joint Vocal-Accompaniment Symbolic Modeling
🤖 AI Summary
本文提出MIDIBack，通过联合建模人声和伴奏事件来解决自动音高修正问题，利用共享的OctupleMIDI序列提高音准准确性。
📝 Abstract
Automatic pitch correction (APC) requires distinguishing the unintended intonation errors from expressive pitch variation. Existing systems either lack explicit harmonic modeling, as vocal-only methods do, or do not directly use the note-level polyphonic context. Therefore, we propose MIDIBack, a note-level APC framework that jointly models the vocal and accompaniment events in a shared OctupleMIDI sequence. We evaluate MIDIBack under 6 note corruption regimes, including global outshift, learned note-dependent detuning, uniform perturbations, and their combinations. The resulting model achieves 78.6% overall raw pitch accuracy (RPA), and 81.5% under combined global outshift and learned detuning. Removing the accompaniment conditioning reduces RPA from 81.5% to 35.8% in outshift, showing the effectiveness of accompaniment context. Case studies on accompaniment modulation further illustrate that vocal note predictions
Innovation
Methods, ideas, or system contributions that make the work stand out.
Joint Vocal-Accompaniment Modeling
OctupleMIDI Sequence
Automatic Pitch Correction (APC)
Harmony-Aware
🔎 Similar Papers
No similar papers found.
