Abstract

Deploying convolutional neural networks (CNNs) on embedded devices is difficult due to the limited memory and computation resources. The redundancy in feature maps is an important characteristic of those successful CNNs, but has rarely been investigated in neural architecture design. This paper proposes a novel Ghost module to generate more feature maps from cheap operations. Based on a set of intrinsic feature maps, we apply a series of linear transformations with cheap cost to generate many ghost feature maps that could fully reveal information underlying intrinsic features. The proposed Ghost module can be taken as a plug-and-play component to upgrade existing convolutional neural networks. Ghost bottlenecks are designed to stack Ghost modules, and then the lightweight GhostNet can be easily established. Experiments conducted on benchmarks demonstrate that the proposed Ghost module is an impressive alternative of convolution layers in baseline models, and our GhostNet can achieve higher recognition performance (e.g. 75.7% top-1 accuracy) than MobileNetV3 with similar computational cost on the ImageNet ILSVRC-2012 classification dataset. Code is available at https://github.com/huawei-noah/ghostnet.

Keywords

Computer scienceRedundancy (engineering)Convolutional neural networkConvolution (computer science)Feature (linguistics)ComputationPattern recognition (psychology)Set (abstract data type)Artificial intelligenceFeature extractionArchitectureComponent (thermodynamics)Data miningArtificial neural networkAlgorithmProgramming language

Affiliated Institutions

Related Publications

Publication Info

Year
2020
Type
article
Pages
1577-1586
Citations
3894
Access
Closed

External Links

Social Impact

Social media, news, blog, policy document mentions

Citation Metrics

3894
OpenAlex

Cite This

Kai Han, Yunhe Wang, Qi Tian et al. (2020). GhostNet: More Features From Cheap Operations. , 1577-1586. https://doi.org/10.1109/cvpr42600.2020.00165

Identifiers

DOI
10.1109/cvpr42600.2020.00165