Parameter Distribution Balanced CNNs

Convolutional neural network (CNN) is the primary technique that has greatly promoted the development of computer vision technologies. However, there is little research on how to allocate parameters in different convolution layers when designing CNNs. We research mainly on revealing the relationship between CNN parameter distribution, i.e., the allocation of parameters in convolution layers, and the discriminative performance of CNN. Unlike previous works, we do not append more elements into the network, such as more convolution layers or denser short connections. We focus on enhancing the discriminative performance of CNN through varying its parameter distribution under strict size constraint. We propose an energy function to represent the CNN parameter distribution, which establishes the connection between the allocation of parameters and the discriminative performance of CNN. Extensive experiments with shallow CNNs on three public image classification data sets demonstrate that the CNN parameter distribution with a higher energy value will promote the model to obtain better performance. According to the motivated observation, the problem of finding the optimal parameter distribution can be transformed into an optimization problem of finding the biggest energy value. We present a simple yet effective guideline that uses balanced parameter distribution to design CNNs. Extensive experiments on ImageNet with three popular backbones, i.e., AlexNet, ResNet34, and ResNet101, demonstrate that the proposed guideline can make consistent improvements upon different baselines under strict size constraint.

I. Introduction

In recent years, deep learning has made great progress in the field of computer vision. Lots of efficient modifications on architectures, regularization techniques, and optimization algorithms promote various computer vision tasks to achieve unprecedented performance, such as image retrieval [1], [2], image labeling [3], object detection and recognition [4], [5], and semantic segmentation [6], [7]. However, as stated in [8], these successes of deep learning owe to the empirical methods, while leaving lots of theoretical puzzles. This is a huge obstacle to the development of deep convolution neural networks. Therefore, this article focuses on revealing the relationship between the CNN parameter distribution, i.e., the allocation of parameters in convolution layers, and the discriminative performance of CNN. Moreover, according to the relationship, we propose a guideline to optimize existing CNNs and design new CNNs to achieve better performance.