Monaural Singing Voice Separation Using Fusion-Net with Time-Frequency Masking | IEEE Conference Publication | IEEE Xplore