Evolutionary multi-objective model compression for deep neural networks

Keyphrases

• Evolutionary multi-objective 100%
• Multi-objective Model 100%
• Model Compression 100%
• Deep Neural Network 100%
• Energy Efficiency 60%
• Energy Model 60%
• Model Size 60%
• Energy Consumption 40%
• Pruning 40%
• Edge Devices 40%
• Face Recognition 20%
• Significant Loss 20%
• Diverse Populations 20%
• Process Optimization 20%
• Orthogonality 20%
• Compression Rate 20%
• Optimization Strategy 20%
• Reduced Time 20%
• Model Accuracy 20%
• Coding Scheme 20%
• Efficient Processing 20%
• Language Translation 20%
• Single Run 20%
• Memory Consumption 20%
• Population Evolution 20%
• Weight Parameter 20%
• VGG16 20%
• Accuracy Loss 20%
• Architecture Search 20%
• CIFAR-10 20%
• Network Pruning 20%
• Dataflow Design 20%
• Neural Network Quantization 20%

Computer Science

• Multiobjective 100%
• Objective Model 100%
• Model Compression 100%
• Deep Neural Network 100%
• Energy Efficiency 60%
• Energy Consumption 40%
• Experimental Result 20%
• Space Complexity 20%
• Process Optimization 20%
• Face Recognition 20%
• Optimization Strategy 20%
• VGG-16 20%
• Compression Rate 20%
• Translation (Languages) 20%

Engineering

• Multi-Objective Model 100%
• Deep Neural Network 100%
• Energy Efficiency 60%
• Experimental Result 20%
• Broad Range 20%
• Orthogonality 20%
• Coding Scheme 20%
• Optimization Strategy 20%