TAIT: One-Shot Full-Integer Lightweight DNN Quantization via Tunable Activation Imbalance Transfer

Keyphrases

• Lightweight Deep Neural Network 100%
• Depthwise Separable Convolution 100%
• Neural Network Quantization 100%
• Integer Quantization 50%
• Parameter Quantization 50%
• Bias Factor 50%
• Framework Design 50%
• Quantization Error 50%
• Quantization Method 50%
• FPGA Acceleration 50%
• Convolution Network 50%
• Edge AI 50%

Computer Science

• Edge AI 16%
• Application Specific Integrated Circuit 16%
• Quantisation Error 16%