Picture for Alessandro Aimar

Alessandro Aimar

NullHop: A Flexible Convolutional Neural Network Accelerator Based on Sparse Representations of Feature Maps

Add code
Mar 06, 2018
Figure 1 for NullHop: A Flexible Convolutional Neural Network Accelerator Based on Sparse Representations of Feature Maps
Figure 2 for NullHop: A Flexible Convolutional Neural Network Accelerator Based on Sparse Representations of Feature Maps
Figure 3 for NullHop: A Flexible Convolutional Neural Network Accelerator Based on Sparse Representations of Feature Maps
Figure 4 for NullHop: A Flexible Convolutional Neural Network Accelerator Based on Sparse Representations of Feature Maps
Viaarxiv icon

ADaPTION: Toolbox and Benchmark for Training Convolutional Neural Networks with Reduced Numerical Precision Weights and Activation

Add code
Nov 13, 2017
Figure 1 for ADaPTION: Toolbox and Benchmark for Training Convolutional Neural Networks with Reduced Numerical Precision Weights and Activation
Figure 2 for ADaPTION: Toolbox and Benchmark for Training Convolutional Neural Networks with Reduced Numerical Precision Weights and Activation
Figure 3 for ADaPTION: Toolbox and Benchmark for Training Convolutional Neural Networks with Reduced Numerical Precision Weights and Activation
Figure 4 for ADaPTION: Toolbox and Benchmark for Training Convolutional Neural Networks with Reduced Numerical Precision Weights and Activation
Viaarxiv icon