A multi-scale fully convolutional network for singing melody extraction

Keyphrases

• Singing Melody Extraction 100%
• Melody Extraction 100%
• Multi-scale Convolutional Network 100%
• Semantic Segmentation 66%
• Encoder 33%
• Decoder 33%
• Less Computation 33%
• Performance Comparison 33%
• Classification Task 33%
• Cepstral Features 33%
• Polyphonic music 33%
• Art Performance 33%
• Audio Processing 33%
• Fully Convolutional Network 33%
• Sequence Task 33%
• Combined Frequency 33%

Engineering

• Experimental Result 100%
• Multiscale 100%
• Input Feature 100%
• Classification Task 100%
• Audio Signal Processing 100%

Computer Science

• Image Segmentation 100%
• Fully Convolutional Network 100%
• cepstral 50%
• Experimental Result 50%
• Classification Task 50%
• Performance Comparison 50%
• Audio Signal Processing 50%
• Art Performance 50%