Reduce Computing Complexity of Deep Neural Networks Through Weight Scaling

Computer Science

• Arithmetic Operation 20%
• Computation Complexity 20%
• Data Reuse 20%
• Deep Neural Network 100%
• Memory Access 20%
• Neural Network Model 60%
• Residual Neural Network 20%
• Sparsity 20%
• VGG-16 20%

Engineering

• Bit Precision 20%
• Computation Complexity 20%
• Deep Neural Network 100%
• Image Recognition 20%
• Memory Access 20%
• Network Model 60%
• Sparsity 20%