TwinDNN: A tale of two deep neural networks

Keyphrases

• Deep Neural Network 100%
• Inference Speed 20%
• Neural Network 10%
• Overall Accuracy 10%
• Model Accuracy 10%
• Image Classification 10%
• Inference Scheme 10%
• Limited Resources 10%
• Model Compression 10%
• Proposed Design 10%
• Resource Availability 10%
• Network Inference 10%
• ImageNet 10%
• Classification Task 10%
• Fast Inference 10%
• Model Size 10%
• Compression Technologies 10%
• Resource Balancing 10%
• CIFAR-10 10%
• Accuracy Degradation 10%
• Network Compression 10%
• Compressed Networks 10%
• VGG16 10%
• Weight Quantization 10%
• Accuracy Recovery 10%

Computer Science

• Deep Neural Network 100%
• Neural Network 10%
• Quantization (Signal Processing) 10%
• Model Accuracy 10%
• Classification Task 10%
• Image Classification 10%
• Network Inference 10%
• Model Compression 10%
• VGG-16 10%