Computational optimization of convolutional neural networks using separated filters architecture
This work addresses efficiency for CNNs in embedded systems, but it is incremental as it builds on existing separable filter techniques.
The paper tackles the computational demands of convolutional neural networks (CNNs) in resource-constrained settings like mobile platforms by proposing a transformation that expresses 2D convolution filters as a linear combination of separable filters, resulting in a 15% speedup without accuracy loss in an industrial image recognition system.
This paper considers a convolutional neural network transformation that reduces computation complexity and thus speedups neural network processing. Usage of convolutional neural networks (CNN) is the standard approach to image recognition despite the fact they can be too computationally demanding, for example for recognition on mobile platforms or in embedded systems. In this paper we propose CNN structure transformation which expresses 2D convolution filters as a linear combination of separable filters. It allows to obtain separated convolutional filters by standard training algorithms. We study the computation efficiency of this structure transformation and suggest fast implementation easily handled by CPU or GPU. We demonstrate that CNNs designed for letter and digit recognition of proposed structure show 15% speedup without accuracy loss in industrial image recognition system. In conclusion, we discuss the question of possible accuracy decrease and the application of proposed transformation to different recognition problems. convolutional neural networks, computational optimization, separable filters, complexity reduction.