Multi-band CNN architecture using adaptive frequency filter for acoustic event classification

Keyphrases

• Classification Performance 100%
• Acoustic Event Classification 100%
• Frequency Filtering 100%
• Adaptive Frequency 100%
• Convolutional Neural Network Architecture 100%
• Spectral Features 66%
• Feature Classification 66%
• High Correlation 66%
• Classification Task 66%
• Filter Parameter 66%
• Training Process 33%
• Neural Network 33%
• Learning Systems 33%
• Goods Classification 33%
• Accurate Classification 33%
• Vulnerability 33%
• Modeling Framework 33%
• Acoustics 33%
• Architecture-centric 33%
• Cut-off Frequency 33%
• Response Characteristics 33%
• Frequency Ratio 33%
• Discriminant Features 33%
• Damping Ratio 33%
• Parameter Modeling 33%
• Filtering Response 33%
• Spectral Bases 33%
• Multi-frequency 33%
• Normalization Method 33%
• Frequency Separation 33%
• Novel Filter 33%
• Stable Feature 33%
• Convolutional Neural Network 33%
• Optimized Frequency 33%
• Learning Parameters 33%
• Filter Frequency Response 33%
• Sub-band Feature 33%
• Neural Network-based 33%
• Frequency Sub-band 33%
• Band Approach 33%
• End-to-end Learning Framework 33%

Computer Science

• Neural Network Architecture 100%
• Convolutional Neural Network 100%
• Classification Performance 75%
• Spectral Feature 50%
• Classification Task 50%
• High Correlation 50%
• Filter Parameter 50%
• Learning Framework 25%
• Training Process 25%
• Modeling Framework 25%
• Neural Network 25%
• Research Effort 25%
• Frequency Response 25%
• Learning Parameter 25%
• Frequency Separation 25%
• Learning System 25%
• Cutoff Frequency 25%

Physics

• Convolutional Neural Network 100%
• Acoustics 100%
• Neural Network 25%
• Cutoff Frequency 25%