Selinet: A Lightweight Model for Single Channel Speech Separation

Keyphrases

• Lightweight Model 100%
• Model Size 100%
• Monaural Speech Separation 100%
• Computational Cost 66%
• Atrous 66%
• Speech Separation 66%
• Bottleneck Module 66%
• Temporal Pyramid Pooling 66%
• Computational Complexity 33%
• Model Performance 33%
• Deep Learning 33%
• Separation Techniques 33%
• State Vector 33%
• Depthwise Separable Convolution 33%
• Computational Model 33%
• Effective Network 33%
• Model Complexity 33%
• Hidden States 33%
• Resource-constrained Devices 33%
• Context Vector 33%
• Squeeze-and-excitation 33%
• Different Fields of View 33%
• Convolutional Architecture 33%

Computer Science

• Computational Cost 100%
• Computational Complexity 50%
• Deep Learning Method 50%