Normalization and dropout for stochastic computing-based deep convolutional neural networks

Keyphrases

• Dropout 100%
• Stochastic Computing 100%
• Deep Convolutional Neural Network (deep CNN) 100%
• Hardware Implementation 28%
• Embedded Systems 28%
• Accuracy Improvement 28%
• Dropout Technique 28%
• Activation Units 28%
• Portable System 14%
• Pattern Recognition 14%
• Training Phase 14%
• Effective Model 14%
• Feature Extraction 14%
• Wearable Devices 14%
• Power Supply 14%
• Limited Power 14%
• Power Consumption 14%
• Recognition Task 14%
• Normalization Method 14%
• Classification Task 14%
• Fast-growing 14%
• Arithmetic Logic Unit 14%
• Complex Topology 14%
• Pattern Classification 14%
• Parallel Counter 14%
• Computing Paradigm 14%
• Limited Area 14%
• Internet of Things Devices 14%
• Dropout Design 14%
• Neural Network Framework 14%
• AlexNet 14%
• Power Complex 14%
• ImageNet Dataset 14%
• Extraction Blocks 14%
• Max Pooling 14%

Computer Science

• Deep Convolutional Neural Networks 100%
• Hardware Implementation 40%
• Power Consumption 20%
• Pattern Classification 20%
• Training Phase 20%
• Internet of Things Device 20%
• computing paradigm 20%
• Wearable Device 20%
• Experimental Result 20%
• Classification Task 20%
• Embedded System 20%
• Feature Extraction 20%
• Pattern Recognition 20%