Hybrid Binary Networks: Optimizing for Accuracy, Efficiency and Memory

Binarization is an extreme network compression approach that provides large computational speedups along with energy and memory savings, albeit at significant accuracy costs. We investigate the question of where to binarize inputs at layer-level granularity and show that selectively binarizing the inputs to specific layers in the network could lead to significant improvements in accuracy while preserving most of the advantages of binarization... (read more)

PDF Abstract

Datasets


Results from the Paper


  Submit results from this paper to get state-of-the-art GitHub badges and help the community compare results to other papers.

Methods used in the Paper


METHOD TYPE
1x1 Convolution
Convolutions
Convolution
Convolutions
Local Response Normalization
Normalization
Grouped Convolution
Convolutions
ReLU
Activation Functions
Dropout
Regularization
Dense Connections
Feedforward Networks
Max Pooling
Pooling Operations
Softmax
Output Functions
AlexNet
Convolutional Neural Networks