Abstract:<p>With the rapid developments of deep neural networks related applications, approaches for accelerating computationally intensive convolutional neural networks, such as network quantization, pruning, knowledge distillation, have attracted ever-increasing attention. Network binarization is an extreme form of network quantization technique, which binarizes the network weights and/or activation values to save computational resources. However, it often introduces noises into the network, and requires larger model size (more parameters) to compensate for the loss of representation capacity. To address the model complexity reduction challenges and further improve the network performance, this paper proposes an approach: slimming binarized neural networks (<em>SBNN</em>), which reduces complexity of binarized networks with acceptable accuracy loss. SBNN prunes the convolutional layers and fully-connected layer in a binarized network. Then it is refined by the proposed SoftSign function, knowledge distillation and full-precision computation to enhance the network accuracy. The proposed SBNN can be also conveniently applied to a pre-trained binarized network. We demonstrate the effectiveness of our approach through several state-of-the-art binarized models. For AlexNet and ResNet-18 on ILSVRC-2012 dataset, SBNN obtains negligible accuracy loss but even a better accuracy than the pre-pruning model while using only 75% of original filters.</p>

Binary Image Thinning Using Autowaves Generated by PCNN

Binary fingerprint image thinning using template-based PCNNs.

Image Thinning Using Pulse Coupled Neural Network

A Neural Network Hardware Model for Image Thinning

A WAVELET-BASED APPROACH TO RIDGE THINNING IN FINGERPRINT IMAGES

Hybrid Parallel Image Processing Algorithm for Binary Images with Image Thinning Technique

A Sequential Thinning Algorithm For Multi-Dimensional Binary Patterns

Binary Image Restoration Using Pulse Coupled Neural Network

A New Method of PCNN's Parameter's Optimization

A novel algorithm of fingerprint thinning based on Gabor phase

Research on Image Processing Based on PCNN

Automatic Image Segmentation Based on PCNN with Adaptive Threshold Time Constant.

Improved Algorithm on Natural Image Matting

Study on node regulation algorithm after thinning

Medical Images Segmentation Based on Improved Three-Dimensional Pulse Coupled Neural Network

A Medial Axis Based Thinning Strategy for Character Images

Image segmentation method based on an improved PCNN

A Background-Thinning Based Algorithm for Separating Connected Handwritten Digit Strings

SBNN: Slimming binarized neural network

A New Approach for Image Shadow Processing Based on PCNN

Dynamical Conventional Neural Network Channel Pruning by Genetic Wavelet Channel Search for Image Classification