Selective Kernel Networks | RDL Research Database

Abstract

In standard Convolutional Neural Networks (CNNs), the receptive fields of artificial neurons in each layer are designed to share the same size. It is well-known in the neuroscience community that the receptive field size of visual cortical neurons are modulated by the stimulus, which has been rarely considered in constructing CNNs. We propose a dynamic selection mechanism in CNNs that allows each neuron to adaptively adjust its receptive field size based on multiple scales of input information. A building block called Selective Kernel (SK) unit is designed, in which multiple branches with different kernel sizes are fused using softmax attention that is guided by the information in these branches. Different attentions on these branches yield different sizes of the effective receptive fields of neurons in the fusion layer. Multiple SK units are stacked to a deep network termed Selective Kernel Networks (SKNets). On the ImageNet and CIFAR benchmarks, we empirically show that SKNet outperforms the existing state-of-the-art architectures with lower model complexity. Detailed analyses show that the neurons in SKNet can capture target objects with different scales, which verifies the capability of neurons for adaptively adjusting their receptive field sizes according to the input. The code and models are available at https://github.com/implus/SKNet.

Keywords

Receptive fieldComputer scienceSoftmax functionArtificial intelligencePattern recognition (psychology)Convolutional neural networkKernel (algebra)Artificial neural networkMathematics

Affiliated Institutions

Related Publications

Network In Network

Min Lin , Qiang Chen , Shuicheng Yan

Abstract: We propose a novel deep network structure called In Network (NIN) to enhance model discriminability for local patches within the receptive field. The conventional con...

2014 arXiv (Cornell University) 1037 citations

Squeeze-and-Excitation Networks

Jie Hu , Li Shen , Samuel Albanie +2 more

The central building block of convolutional neural networks (CNNs) is the convolution operator, which enables networks to construct informative features by fusing both spatial a...

2019 IEEE Transactions on Pattern Analysis... 12023 citations

Conditional Random Fields as Recurrent Neural Networks

Shuai Zheng , Sadeep Jayasumana , Bernardino Romera‐Paredes +5 more

Pixel-level labelling tasks, such as semantic segmentation, play a central role in image understanding. Recent approaches have attempted to harness the capabilities of deep lear...

2015 2381 citations

Fully convolutional networks for semantic segmentation

Jonathan Long , Evan Shelhamer , Trevor Darrell

Convolutional networks are powerful visual models that yield hierarchies of features. We show that convolutional networks by themselves, trained end-to-end, pixels-to-pixels, ex...

2015 35498 citations

ECA-Net: Efficient Channel Attention for Deep Convolutional Neural Networks

Qilong Wang , Banggu Wu , Pengfei Zhu +3 more

Recently, channel attention mechanism has demonstrated to offer great potential in improving the performance of deep convolutional neural networks (CNNs). However, most existing...

2020 6942 citations

Publication Info

Year: 2019
Type: article
Pages: 510-519
Citations: 2769
Access: Closed

External Links

View on DOI.org

Social Impact

Altmetric

Selective Kernel Networks

PlumX Metrics

Social media, news, blog, policy document mentions

Citation Metrics

2769

OpenAlex

Cite This

APA Style

                            
                                    Xiang Li, 
                                
                                    Wenhai Wang, 
                                
                                    Xiaolin Hu
                                
                                et al.
                            
                            (2019). 
                            Selective Kernel Networks. 
                            
                            , 510-519.
                            https://doi.org/10.1109/cvpr.2019.00060

Identifiers

DOI: 10.1109/cvpr.2019.00060