Boosting Convolutional Neural Networks with Middle Spectrum Grouped Convolution

by   Zhuo Su, et al.

This paper proposes a novel module called middle spectrum grouped convolution (MSGC) for efficient deep convolutional neural networks (DCNNs) with the mechanism of grouped convolution. It explores the broad "middle spectrum" area between channel pruning and conventional grouped convolution. Compared with channel pruning, MSGC can retain most of the information from the input feature maps due to the group mechanism; compared with grouped convolution, MSGC benefits from the learnability, the core of channel pruning, for constructing its group topology, leading to better channel division. The middle spectrum area is unfolded along four dimensions: group-wise, layer-wise, sample-wise, and attention-wise, making it possible to reveal more powerful and interpretable structures. As a result, the proposed module acts as a booster that can reduce the computational cost of the host backbones for general image recognition with even improved predictive accuracy. For example, in the experiments on ImageNet dataset for image classification, MSGC can reduce the multiply-accumulates (MACs) of ResNet-18 and ResNet-50 by half but still increase the Top-1 accuracy by more than 1 can also increase the Top-1 accuracy of the MobileNetV2 backbone. Results on MS COCO dataset for object detection show similar observations. Our code and trained models are available at


page 1

page 4

page 5

page 9

page 10


An Attention Module for Convolutional Neural Networks

Attention mechanism has been regarded as an advanced technique to captur...

EPSANet: An Efficient Pyramid Split Attention Block on Convolutional Neural Network

Recently, it has been demonstrated that the performance of a deep convol...

Contextual Convolutional Neural Networks

We propose contextual convolution (CoConv) for visual recognition. CoCon...

A closer look at network resolution for efficient network design

There is growing interest in designing lightweight neural networks for m...

Wise-SrNet: A Novel Architecture for Enhancing Image Classification by Learning Spatial Resolution of Feature Maps

One of the main challenges since the advancement of convolutional neural...

Spatial Group-wise Enhance: Improving Semantic Feature Learning in Convolutional Networks

The Convolutional Neural Networks (CNNs) generate the feature representa...

Dynamic Runtime Feature Map Pruning

High bandwidth requirements are an obstacle for accelerating the trainin...

Please sign up or login with your details

Forgot password? Click here to reset