A privacy-preserving-oriented DNN pruning and mobile acceleration framework

Keyphrases

• Pruning Strategy 100%
• Privacy-preserving 100%
• Pruning 100%
• Acceleration Framework 100%
• Weight Pruning 100%
• Deep Neural Network Pruning 100%
• Mobile Acceleration 100%
• Performance Improvement 50%
• Deep Neural Network 50%
• Non-expert 50%
• Compiler 50%
• Alternating Direction multiplier Method 50%
• Synthetic Data 50%
• Training Data 50%
• Limited Storage Capacity 50%
• User Data 50%
• Computing Capabilities 50%
• Data Privacy 50%
• Model Size 50%
• Storage Capability 50%
• Accuracy Loss 50%
• Algorithmic Level 50%
• Neural Network Framework 50%
• End-to-end Deep Learning 50%
• Private Training 50%
• Inference Acceleration 50%
• Mobile Edge Devices 50%
• TensorFlow Lite 50%
• Pattern Pruning 50%

Computer Science

• Deep Neural Network 100%
• Privacy Preserving 100%
• User Data 33%
• Data Privacy 33%
• Alternating Direction Method of Multipliers 33%
• Experimental Result 33%
• Synthetic Data 33%
• Training Dataset 33%
• TensorFlow 33%