Quantized convolutional neural networks through the lens of partial differential equations

Keyphrases

• Convolutional Neural Network 100%
• Partial Differential Equations 100%
• Low Power 20%
• Neural Network 20%
• Total Variation 20%
• Computational Burden 20%
• Low-resource 20%
• Accuracy Improvement 20%
• Time Application 20%
• PDE 20%
• Image Classification 20%
• Piecewise Constant 20%
• Resource-constrained 20%
• Value Distribution 20%
• Sensitive Resource 20%
• Edge-preserving Smoothing 20%
• Autonomous Driving 20%
• Time Stability 20%
• Variation Approach 20%
• Fixed-point Arithmetic 20%
• Feature Map 20%
• Edge Devices 20%
• Graph Convolutional Network 20%
• Graph Node Classification 20%
• Forward Stability 20%
• Quantized Networks 20%
• Image Graph 20%

Computer Science

• Partial Differential Equation 100%
• Convolutional Neural Network 100%
• Quantization (Signal Processing) 60%
• Neural Network 20%
• Image Classification 20%
• Real-Time Application 20%
• Total Variation 20%
• fixed-point arithmetic 20%
• Autonomous Driving 20%
• Graph Convolutional Network 20%
• Node Classification 20%

Mathematics

• Partial Differential Equation 100%
• Convolutional Neural Network 100%
• Edge 40%
• Total Variation 20%
• Fixed Points 20%
• Low Power 20%
• Arithmetic 20%
• Neural Network 20%