Sparse kernel cepstral coefficients (SKCC): Inner-product based features for noise-robust speech recognition

Keyphrases

• Product Basis 100%
• End-to-end Speech Recognition 100%
• Inner Product 100%
• Sparse Kernel 100%
• Cepstral Coefficients 100%
• Speech Signal 40%
• Speech Recognition 20%
• Recording Conditions 20%
• Feature Extraction Methods 20%
• Noise-robust 20%
• Recognition Performance 20%
• Reproducing Kernel Hilbert Space 20%
• Encoding Method 20%
• Regression Techniques 20%
• Aurora 20%
• Art Features 20%
• Speech Feature Extraction 20%
• Over-complete 20%
• Robust Information 20%
• DSP-based 20%
• Noise Robustness 20%
• Auditory Coding 20%
• Sparsification 20%

Computer Science

• Robust Speech Recognition 100%
• Cepstral Coefficient 100%
• Speech Recognition 20%
• Reproducing Kernel 20%
• Hilbert Space 20%
• Basis Function 20%
• Recognition Performance 20%
• Feature Extraction 20%