CNN Based Two-stage Multi-resolution End-to-end Model for Singing Melody Extraction.

Ming-Tso Chen, Bo-Jun Li, Tai-Shih Chi

Browse the full ICASSP paper archive.