Constructing Fast Network through Deconstruction of Convolution

Abstract

Convolutional neural networks have achieved great success in various vision tasks; however, they incur heavy resource costs. By using deeper and wider networks, network accuracy can be improved rapidly. However, in an environment with limited resources (e.g., mobile applications), heavy networks may not be usable. This study shows that naive convolution can be deconstructed into a shift operation and pointwise convolution. To cope with various convolutions, we propose a new shift operation called active shift layer (ASL) that formulates the amount of shift as a learnable function with shift parameters. This new layer can be optimized end-to-end through backpropagation and it can provide optimal shift values. Finally, we apply this layer to a light and fast network that surpasses existing state-of-the-art networks.

arXiv commentTo be appeared in NIPS 2018

References in corpus (5)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
• SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size
• Caffe: Convolutional Architecture for Fast Feature Embedding
• Deformable Convolutional Networks

Cited by in corpus (3)

• GhostNets on Heterogeneous Devices via Cheap Operations
• Learning Versatile Convolution Filters for Efficient Visual Recognition
• Full-Stack Filters to Build Minimum Viable CNNs