Efficient memory compression in deep neural networks using coarse-grain sparsification for speech applications

Computer Science

• Deep Neural Network 100%
• Speech Application 100%
• Speech Recognition 33%
• Neural Network 16%
• Wearable Device 16%
• Mobile Device 16%
• Memory Footprint 16%
• Low-Power Design 16%
• Sparsity 16%
• Detection Accuracy 16%
• Network Layer 16%
• Network Design 16%
• Recognition Network 16%
• Neural Network Training 16%
• Resource Requirement 16%
• Word Error Rate 16%
• Memory Requirement 16%
• Double Precision 16%

Engineering

• Deep Neural Network 100%
• Speech Application 100%
• Error Rate 33%
• Wearable Sensor 16%
• Memory Footprint 16%
• Sparsity 16%
• Real Hardware 16%
• Neural Network Training 16%
• Memory Requirement 16%
• Layer Network 16%
• Double Precision 16%

Keyphrases

• Deep Neural Network Design 25%