Dual stage learning based dynamic time-frequency mask generation for audio event classification

Keyphrases

• Learning-based 100%
• Dynamic Time 100%
• Time-frequency Mask 100%
• Mask Generation 100%
• Dual-stage 100%
• Audio Event Classification 100%
• Feature Enhancement 66%
• Noisy Environment 66%
• Dynamic Filter 66%
• Generator Network 66%
• Performance Improvement 33%
• Audio 33%
• High-noise Environments 33%
• Enhancement Mechanism 33%
• Noise Background 33%
• Noise Problem 33%
• Mask Based 33%
• Filter Set 33%
• Event Recognition 33%
• Unknown Noise 33%
• Unseen Noise 33%
• Filter Generator 33%
• Fixed Filters 33%
• Audio-based 33%

Computer Science

• Generator Network 100%
• Audio Event Classification 100%
• Background Noise 50%
• Noisy Environment 50%

Engineering

• Filter Dynamic 100%
• Noisy Environment 50%