Harmonic-aware tri-path convolution recurrent network for singing voice separation

Keyphrases

• Singing Voice Separation 100%
• Dual-path 100%
• Recurrent Network 100%
• Temporal Coherence 66%
• Recurrent Network Model 66%
• Interband 33%
• Evaluation Results 33%
• Spectrogram 33%
• Public Dataset 33%
• Human Auditory System 33%
• Separation Performance 33%
• Temporal Dimension 33%
• Spectral Dimension 33%
• Streaming Process 33%
• Separation Model 33%
• Sound Separation 33%
• Spectral Regularity 33%
• Auditory Streaming 33%
• Recurring Neural Networks 33%
• Conv-TasNet 33%

Computer Science

• Recurrent Network 100%
• Recurrent Neural Network 20%
• Neural Network 20%
• Evaluation Result 20%
• General Pattern 20%
• Temporal Dimension 20%
• Spectral Dimension 20%
• Separation Performance 20%